While influence operations research has traditionally focused on high-arousal emotions like anger and outrage, this report by Alina Bârgăoanu and Maya Sobchuk examines a quieter, more insidious vector: nostalgia. Unlike overt propaganda, weaponized nostalgia operates as an "ambient" emotion, it doesn't announce itself as political, but instead shapes how audiences judge the present through idealized, emotionally warm portrayals of the past, gradually normalizing distrust in current institutions and eroding critical evaluation. Drawing on two original investigations, the report documents this dynamic in action. In Romania, a Russian-linked network of roughly 75 Facebook pages and affiliated clickbait sites active since 2020 and reaching a combined audience in the millions has spent years embedding nostalgia for the communist era across 12,790 analyzed posts. Rather than explicit political messaging, the content leans on themes of family, faith, childhood, tradition, and "true" values, implicitly contrasting a morally coherent past with a fragmented present. Notably, these pages were built years in advance of any electoral target and have survived repeated exposure, including through the contested 2024 election cycle. In Okinawa, Japan, the report documents PRC-aligned campaigns exploiting Ryukyuan historical identity across X/Twitter, YouTube, and Facebook. These efforts range from official Chinese state media (including a dedicated Global Times hashtag, #RyukyuChronicles) to coordinated networks of inauthentic accounts some AI-generated, some sloppily assembled with mismatched VPNs and bios promoting narratives of Ryukyu's "undetermined status," historical grievance against Japan, and opposition to U.S. military presence. Over 44–56% of accounts engaging in these narratives displayed bot-like characteristics. The comparative analysis reveals a shared playbook: nostalgic content is dressed up as cultural celebration, historical education, or community memory, making it resistant to being dismissed as propaganda. The report also identifies a complementary strategy, pairing nostalgic narratives with "techno-utopian" messaging (as seen in Chinese diplomatic content in Malaysia) that frames a China-aligned future as the answer to a present rendered inadequate from both temporal directions. The authors argue that because nostalgia is ambient rather than viral, it evades most automated detection systems, making manual, qualitative analysis essential. The report closes with recommendations for influence defense practitioners: engage the emotional register of these narratives rather than relying on factual rebuttal alone, build cognitive resilience before narratives normalize, and treat slow narrative permeation as seriously as sudden disinformation bursts. Authors: Alina Bârgăoanu & Maya Sobchuk [Download PDF Here]

Indonesia's June 2026 protest wave, driven by university students, women's groups, and civic actors reacting to economic pressure, fuel and food costs, and criticism of the Free Nutritious Meals Programme was locally rooted, but its meaning was quickly contested online. This report documents how a cluster of pro-PRC influencer accounts moved to reframe the demonstrations as a U.S.-, NED-, and Soros-backed "color revolution" aimed at destabilizing Indonesia and countering China's regional influence. Using narrative intelligence tools, CRC researchers identified 187 relevant posts generating over 1.2 million views, concentrated in two coordinated surges (June 6–7 and June 14–15) that tracked closely with key moments in the protest cycle. Four accounts @angeloinchina, @NuryVittachi, @BrianJBerletic, and @PeterCronau drove the bulk of this activity, mutually amplifying one another and receiving support from a network of 37 accounts, 45% of which displayed bot-like behavior. Notably, this was not an isolated incident: the same accounts pushed similar claims during Indonesia's 2025 protest cycle and timed a parallel revival of "Tiananmen was Western-backed" narratives to coincide with the June 4th memorial period, pointing to a recurring, deliberate playbook rather than spontaneous commentary. While open-source evidence does not support claims that the protests were foreign-made or externally funded, the report argues that Indonesia as Southeast Asia's largest economy and a longtime adherent of a non-aligned foreign policy, remains a key target for narratives seeking to shape public and policymaker perceptions amid intensifying U.S.-China strategic competition. [Download PDF Here]

Commissioned by the Friedrich Naumann Foundation For Freedom Malaysia in late 2025, as part of a Study on Perceptions of Geopolitics and Regional Issues, this dedicated CRC report examines how Chinese and Russian influence affects Malaysia’s digital information environment and public opinion. The report integrates the results of two recent national surveys, narrative intelligence findings, and a media environment analysis. It argues, that Malaysia is not necessarily undergoing authoritarian conversion. Instead, it is consistently exposed to foreign-driven anti-Western and authoritarian-aligned narratives which are embedded across state media, diplomatic channels, local outlets, social media platforms, and amplifier assets. In that context, the People's Republic of China (PRC) is the most prominent foreign actor, using a multi-layered influence architecture in an attempt to frame Beijing as Malaysia’s foremost economic and strategic partner. At the same time, Russian influence activity is also observed, albeit on a more limited scale. It relies mostly on diplomatic messaging, cultural institutions, individual influencers, and media partnerships promoting narratives around multipolarity, sovereignty, anti-Western sentiment, and closer Malaysia-Russia alignment. Survey data from the CRC and Merdeka Center provides important insights. It shows that although many Malaysians view the PRC as highly active and economically beneficial, a majority still holds China responsible for South China Sea tensions. For Malaysian and European influence defense stakeholders, the report highlights uneven cognitive resilience capacity across demographic groups, reinforcing the need for improved strategic communication deployment, as well as adoption of counter-FIMI detection, remediation and response capabilities in order to protect the crucial assets of democracy. This report was compiled in March 2026. [Download PDF Here]

Since the outbreak of the Iran War (Operation Epic Fury) on 28 February 2026, the conflict has emerged as a landmark case study in modern hybrid warfare, one defined not just by airstrikes and military force, but by the seamless integration of cyber operations and information warfare into a unified offensive strategy. Cyfluence Research Center (CRC) examines what analysts are calling "cyfluence" operations: the coordinated fusion of cyber capabilities with influence campaigns designed to shape perceptions, sow confusion, and degrade morale. Key incidents documented include the compromise of Iran's BadeSaba prayer app, which was hijacked on the first day of strikes to push surrender messages to millions of Iranian users, and the simultaneous kinetic strike and broadcast hijack of state television network IRIB, through which messages from Israeli PM Netanyahu and President Trump were beamed directly to Iranian audiences. Beyond these high-profile operations, the report details Iranian-linked disinformation efforts, including networks of sockpuppet accounts impersonating Chinese, Russian, and North Korean state media, alongside the prolific hack-and-leak activities of the Iran-aligned hacktivist group Handala, whose claimed targets ranged from Israeli universities to FBI Director Kash Patel's personal email. The report concludes that cyfluence is no longer a peripheral tactic, it has become the operational logic of modern warfare, where controlling how events are perceived may matter as much as controlling territory. Key Takeaways The current war in Iran and the wider Middle East (also known as Operation Epic Fury and Operation Roaring Lion) has demonstrated a remarkably close integration between varied forms of kinetic and hybrid warfare, including combined cyber-enabled influence (cyfluence) operations, employed in tandem to maximize strategic effects. As such, Hostile Influence Campaigns (HICs) and Coordinated Information Disorder (CID) increasingly function as Primary Offensive Efforts (POEs) alongside the use of military force. Russia and China – Iran’s most powerful allies – have played an important role during the conflict, providing continuous diplomatic backing, supplying tactical intelligence, and executing supportive offensive information operations, while exploiting newly-created cognitive attack surfaces. AI-assisted DISARM mapping is employed throughout this report to render cyfluence attack chains analytically tractable, building on methodologies for agentic AI operationalization of influence operation analysis frameworks. Author: The CRC Team [Download PDF Here]

[Introduction] Cyber-based hostile influence campaigns are aimed at influencing target audiences by promoting information and/or disinformation over the internet, sometimes combined with cyber-attacks which enhance their effect (hence force Cyfluence, as opposed to cyber-attacks that aim to steal information, extort money, etc.) Such hostile influence campaigns and operations can be considered an epistemological branch of Information Operations (IO) or Information Warfare (IW). Typically, and as customary during the last decade, the information is spread throughout various internet platforms, which are the different elements of the hostile influence campaign, and as such, connectivity and repetitiveness of content between several elements are the main core characteristics of influence campaigns. Hostile influence campaigns, much like Cyber-attacks, have also become a tool for rival nations and corporations to damage reputation or achieve various business, political or ideological goals. Much like in the cyber security arena, PR professionals and government agencies are responding to negative publicity and disinformation shared over the news and social media. We use the term cyber based hostile influence campaigns, as we include in this definition also cyber-attacks aimed at influencing (such as hack and leak during election time), while we exclude of this term other types of more traditional kinds of influence such as diplomatic, economic, military etc. During the 10th to 16th August 2026, we observed, collected and analyzed endpoints of information related to cyber based hostile influence campaigns (including Cyfluence attacks). The following report is a summary of what we regard as the main events. Some of the mentioned campaigns have to do with social media and news outlets solemnly, while others leverage cyber-attack capabilities. [Contents] [Introduction] [Report Highlights] [Report Summary] [State Actors] Russia Operation Matryoshka Deploys Fake Websites to Target Germany's 2026 Election Russian Fabricated Nova Poshta Strikes, Matryoshka France Deepfakes, and Invented British Mercenary Russia's GRU Training Camps Targeted Moldova Matryoshka Pushes Fake Reports to Exploit Germany's East-West Divide Russian Fabrication Claims Ukrainian Drone Damaged Portuguese F-16s Russian Manipulation False-Flag Allegation at Leipzig Airport China Chinese AI Chatbots Fail to Debunk Pro-China False Claims China's PLA Deploys GoLaxy Smart Propaganda System Against Taiwan [AI Related Articles] Grok Produced 87% of Synthetic Audio Files South Korea's KNPA Arrests people for Non-Consensual Deepfake Sexual Content AI Detection Tools Correctly Identified Fewer AI-Generated Election Images Facebook AI-Generated Silver Fox Personas Accumulate Up to One Million Views [General Reports] Online Anti-Migrant Narratives on X, Telegram, and TikTok Preceded Violence in Southampton and Belfast Michigan Local Election Officials Anticipate Significant Disinformation Problems Ahead of November 2026 Midterms Brazil Orders Discord to Suspend Livestreaming Voter Registration and Biometric Verification Systems Pose Greater Structural Risk Than Deepfakes [Appendix - Frameworks to Counter Disinformation] Nationwide Partnership Network to Fact-Check Critical 2026 Midterm Elections FTC Proposes Regulating Undisclosed AI Ideological Bias [CRC Glossary] [ Report Highlights] EUObserver reported that Operation Matryoshka, a Russia-linked influence network, is deploying websites impersonating BBC News and German public broadcaster ARD to spread disinformation targeting Germany's September 2026 federal election, applying the same impersonation infrastructure previously used against French audiences during the May 2026 regional elections. An NYT investigation based on intelligence officials from three countries, Moldovan prosecutors, and leaked documents reveals that Russia's Kremlin ran GRU-linked paramilitary training camps in Serbia, Bosnia, and Moscow, paid hundreds of Orthodox priests to push Kremlin narratives, making Moldova the most extensively documented Russian election interference operation outside Ukraine. NewsGuard documented Operation Matryoshka's campaign to exploit Germany's historical East-West divide ahead of state elections, using the hashtag #TimeToDivideGermany, and designed primarily to seed false claims into AI tools and fact-checking ecosystems rather than generate direct website traffic. FPRI analyzed China's cognitive warfare playbook against Taiwan, documenting the GoLaxy Smart Propaganda System, a PLA AI platform automating content generation, multi-platform distribution, and audience targeting, as part of a broader PLA cognitive domain operations doctrine that treats information warfare as an integrated component of military operations. Resemble AI's H1 2026 Deepfake Threat Report identified 15,736 confirmed victims of deepfake audio and video attacks globally in the first six months of 2026, with Grok accounting for 87% of detected synthetic audio files, with financial fraud, non-consensual intimate imagery, and political impersonation the three dominant use categories. ISD Global traced the amplification pathway from false and misleading anti-migrant posts on X, Telegram, and TikTok to real-world violence at asylum seeker accommodation in Southampton and Belfast, indicating that platform enforcement actions taken after 2024 have not disrupted the underlying far-right network infrastructure. A RestOfWorld analysis by the International Foundation for Electoral Systems and the African Union's Advisory Group on AI argued that disproportionate focus on deepfakes understates the structural risk of AI embedded in voter registration, biometric verification, and results management systems. PolitiFact announced a nationwide partnership network combining state and local news organizations, university journalism programs, and digital publishers to extend coordinated fact-checking coverage to smaller markets and competitive congressional districts historically underserved by national fact-checkers ahead of the November 2026 US midterm elections. [ Report Summary] StopFake debunked a Russian propaganda claim that a Ukrainian drone damaged Portuguese F-16 aircraft stationed at Estonia's Amari Air Base, a fabrication circulating across pro-Russian Telegram channels. StopFake documented how Russian propaganda channels stripped caveats from Die Zeit reporting on an incident at Leipzig/Halle Airport to falsely present the German newspaper as concluding Ukraine staged a false-flag operation, when Die Zeit's actual reporting raised questions without making any such allegation. EUObserver reported that Operation Matryoshka, a Russia-linked influence network, is deploying websites impersonating BBC News and German public broadcaster ARD to spread disinformation targeting Germany's September 2026 federal election, applying the same impersonation infrastructure previously used against French audiences during the May 2026 regional elections. Detector Media's weekly review of Russian disinformation for 5-11 August 2026 documented fabricated claims of Russian strikes on the Nova Poshta postal service, Matryoshka network deepfake content targeting French audiences, and an entirely invented story about a captured British mercenary in Ukraine, each serving distinct narrative goals across different geographic target audiences. An NYT investigation based on intelligence officials from three countries, Moldovan prosecutors, and leaked documents reveals that Russia's Kremlin ran GRU-linked paramilitary training camps in Serbia, Bosnia, and Moscow, paid hundreds of Orthodox priests to push Kremlin narratives, making Moldova the most extensively documented Russian election interference operation outside Ukraine. NewsGuard documented Operation Matryoshka's campaign to exploit Germany's historical East-West divide ahead of state elections, using the hashtag #TimeToDivideGermany, and designed primarily to seed false claims into AI tools and fact-checking ecosystems rather than generate direct website traffic. NewsGuard tested Chinese-made AI chatbots including DeepSeek, Kimi, and Doubao against non-Chinese models on the same pro-China false claims, finding Chinese chatbots failed to debunk those claims 53% of the time versus 24% for non-Chinese models. FPRI analyzed China's cognitive warfare playbook against Taiwan, documenting the GoLaxy Smart Propaganda System, a PLA AI platform automating content generation, multi-platform distribution, and audience targeting, as part of a broader PLA cognitive domain operations doctrine that treats information warfare as an integrated component of military operations. Resemble AI's H1 2026 Deepfake Threat Report identified 15,736 confirmed victims of deepfake audio and video attacks globally in the first six months of 2026, with Grok accounting for 87% of detected synthetic audio files, with financial fraud, non-consensual intimate imagery, and political impersonation the three dominant use categories. South Korea's National Police Agency arrested 419 people in 2026 for producing or distributing non-consensual AI-generated deepfake sexual content, primarily male students targeting female classmates, while deploying a physics-based AI detection system that identifies synthetic images through lighting inconsistencies and facial geometry anomalies that current AI generators cannot fully simulate. The Brennan Center for Justice found that AI detection tools tested on AI-generated election-related images correctly identified fewer synthetic images on average, demonstrating that both technological and human-review safeguards for AI-generated election content are operating below reliable identification thresholds ahead of the November 2026 US midterms. Futurism documented networks of AI-generated fake male personas on Facebook monetized through Facebook's own Content Monetization program, with Google's AI Overview initially presenting the fabricated identities as real people when queried by users. ISD Global traced the amplification pathway from false and misleading anti-migrant posts on X, Telegram, and TikTok to real-world violence at asylum seeker accommodation in Southampton and Belfast, indicating that platform enforcement actions taken after 2024 have not disrupted the underlying far-right network infrastructure. A survey of Michigan local election officials found 80% confident in voting integrity ahead of the November 2026 midterms, while 20% anticipated significant disinformation problems, with the officials concerned disproportionately representing smaller rural jurisdictions. Brazil's Supreme Court ordered Discord to suspend its livestreaming functionality after the platform repeatedly failed to comply with government orders to remove neo-Nazi content, extremist channels, and material promoting violence against minority groups, representing an escalation from individual content takedowns to targeted functional suspensions as a compliance pressure mechanism. A RestOfWorld analysis by the International Foundation for Electoral Systems and the African Union's Advisory Group on AI argued that disproportionate focus on deepfakes understates the structural risk of AI embedded in voter registration, biometric verification, and results management systems. PolitiFact announced a nationwide partnership network combining state and local news organizations, university journalism programs, and digital publishers to extend coordinated fact-checking coverage to smaller markets and competitive congressional districts historically underserved by national fact-checkers ahead of the November 2026 US midterm elections. The Federal Trade Commission proposed applying Section 5 of the FTC Act, which prohibits unfair or deceptive acts or practices, to AI systems that embed or amplify ideological bias in their outputs without disclosing that bias to users, advancing a theory of liability in which undisclosed systematic political skew in AI responses constitutes a deceptive trade practice subject to FTC enforcement. [State Actors] Russia Operation Matryoshka Deploys Fake Websites to Target Germany's 2026 Election An investigation published by EUObserver states that Operation Matryoshka, a Russia-linked influence network, is deploying websites impersonating BBC News and German public broadcaster ARD to spread disinformation targeting Germany's September 2026 federal election, with fabricated content designed to undermine confidence in German democratic institutions and amplify divisive narratives by routing them through the reputational authority of trusted international and domestic broadcasters. The investigation states that Operation Matryoshka's targeting of Germany's federal election represents systematic adaptation of proven infrastructure across sequential European electoral cycles; the same BBC and broadcaster-impersonation methods previously deployed against French audiences during May 2026 regional elections are now being applied to Germany's federal vote, indicating an operational doctrine of reusing tested influence infrastructure rather than developing bespoke campaigns for each national target. Source: EUobserver. Operation Matryoshka: Russia using ‘BBC news’ fakes to pervert key German vote. [online] Published 11 August 2026. Available at: https://euobserver.com/231951/operation-matryoshka-russia-using-bbc-news-fakes-to-pervert-key-german-vote/ Top Of Page Russian Fabricated Nova Poshta Strikes, Matryoshka France Deepfakes, and Invented British Mercenary A report published by Detector Media states that Russian disinformation during 5th to 11th August 2026 included fabricated claims of Russian strikes on Nova Poshta, Ukraine's largest private postal and logistics company, designed to demoralize the Ukrainian civilian population by manufacturing evidence of logistical collapse, alongside Matryoshka network deepfake content targeting French audiences and an entirely invented story about a captured British mercenary fighting in Ukraine. The report states that the week's three simultaneous fabrication tracks illustrate the Kremlin's multi-audience targeting doctrine: the Nova Poshta narrative targets Ukrainian civilian morale, the Matryoshka deepfakes target French electoral opinion ahead of Germany's September vote, and the British mercenary fabrication sustains the false NATO-involvement framing used in Russian domestic propaganda and Global South channels to justify escalation, three geographically distinct audiences served by three operationally distinct but simultaneously active fabrication tracks. Source: Detector Media. Strikes on Nova Poshta and a fabricated “British mercenary.” Review of Russian fakes from August 5–11, 2026. [online] Published 13 August 2026. Available at: https://en.detector.media/post/strikes-on-nova-poshta-and-a-fabricated-british-mercenary-review-of-russian-fakes-from-august-5-11-2026 Top Of Page Russia's GRU Training Camps Targeted Moldova An investigation published by The New York Times states that Russia conducted one of the most extensive foreign election interference operations documented outside Ukraine against Moldova, deploying GRU-linked training camps in Serbia, Bosnia, and Moscow to build paramilitary shock troops; paying hundreds of Orthodox priests approximately USD 1,000 each to campaign against EU membership from the pulpit; and funding exiled oligarch Ilan Shor's Skolkovo-based vote-buying scheme, which enlisted over 150,000 Moldovans through 240 Telegram chatbots and paid approximately USD 20 million to 38,000 activists ahead of Moldova's 2024 EU membership referendum, according to leaked documents, three Western intelligence services, and Moldovan prosecutors. According to the investigation, Putin privately told military commanders in December 2025 that Russia's strategic goal for 2026 is 'the collapse of NATO and the E.U. from within, with Moldova ranked second only to Ukraine in Kremlin priorities, both because it sits between NATO members and Ukraine and because Russia still seeks to create a land bridge through southern Ukraine to Transnistria, the Russian-controlled breakaway region inside Moldova, making the documented scale of covert investment in Moldovan electoral politics a strategic imperative rather than a peripheral influence campaign, and one whose methods (GRU paramilitary training, priest networks, chatbot-coordinated vote-buying, CEC hacking) represent a composite template for Russian electoral subversion that Western intelligence expects to be adapted for future targets. Source: The New York Times. Putin’s Russia Seeks to Destabilize Moldova. [online] Published 14 August 2026. Available at: https://www.nytimes.com/2026/08/14/world/europe/russia-moldova-putin.html Top Of Page Matryoshka Pushes Fake Reports to Exploit Germany's East-West Divide An investigation published by NewsGuard states that Operation Matryoshka has produced at least 20 fabricated videos and news reports depicting German news outlets, celebrities, and business leaders as advocates for the redivision of Germany into its communist-era East and West, circulated under the hashtag #TimeToDivideGermany ahead of Germany's state elections, with the broader Matryoshka campaign comprising 723 fake news reports mimicking 134 real news brands, including Der Spiegel and Deutsche Welle among the most imitated, and designed primarily not to generate traffic to fake websites but to seed false claims that propagate through organic repetition, harm credible media reputations, overwhelm fact-checkers, and manipulate AI language models that synthesize across sources. The investigation states that Operation Matryoshka's German redivision campaign represents a documented strategic inversion in Russian messaging about Germany, shifting from the prior Kremlin position of crediting Russia for Germany's peaceful reunification to actively attempting to reverse it, a shift attributable to Germany's sustained military and financial support for Ukraine, and one that exploits residual East-West socioeconomic tensions as a social fault line to fracture the domestic political consensus underpinning Germany's Ukraine policy, with the redivision framing calibrated to audiences in the former East Germany where support for continued Ukraine assistance has historically been lower. Source: NewsGuard. Russia’s Push to Re-Divide Germany. [online] Available at: https://www.newsguardrealitycheck.com/p/russias-push-to-re-divide-germany Top Of Page Russian Fabrication Claims Ukrainian Drone Damaged Portuguese F-16s An analysis published by StopFake states that Russian propaganda spread a false claim that a Ukrainian drone struck and damaged Portuguese F-16 aircraft stationed at Estonia's Amari Air Base, a fabrication circulating across pro-Russian Telegram channels and websites without any corroborating evidence and contradicted by both Estonian and Portuguese defence officials who confirmed no such incident occurred. The analysis states that the fabricated claim follows an established Kremlin pattern of inventing drone incidents at NATO facilities to portray alliance member states as direct participants in the Ukraine conflict, with the specific targeting of Portuguese aircraft at an Estonian base designed to simultaneously strain Portugal-Estonia relations and amplify narratives of NATO complicity in Ukrainian strikes against Russia. Source: StopFake. Fake: Ukrainian Drone Damaged Portuguese F-16s at Estonian Air Base. [online] Published 15 August 2026. Available at: https://www.stopfake.org/en/fake-ukrainian-drone-damaged-portuguese-f-16s-at-estonian-air-base/ Top Of Page Russian Manipulation False-Flag Allegation at Leipzig Airport An analysis published by StopFake states that Russian propaganda channels misrepresented reporting by the German newspaper Die Zeit to falsely claim the outlet concluded Ukraine staged a false-flag operation at Leipzig/Halle Airport, when Die Zeit's actual reporting raised questions about an incident without making any such allegation, with Russian sources stripping all qualifiers and caveats to convert acknowledged journalistic uncertainty into a definitive claim of Ukrainian culpability. The analysis states that the manipulation follows a documented Kremlin technique of selectively quoting or misrepresenting credible Western media to lend false authority to disinformation narratives, exploiting the reputational capital of established outlets like Die Zeit to reach audiences that would reject the same claim when it originates from Russian state media. This technique makes the false narrative self-insulating against debunking, since the cited outlet genuinely exists and did publish the source article. Source: StopFake. Manipulation: Die Zeit Says Ukraine May Have Staged a False-Flag Operation at Leipzig Airport. [online] Published 14 August 2026. Available at: https://www.stopfake.org/en/manipulation-die-zeit-says-ukraine-may-have-staged-a-false-flag-operation-at-leipzig-airport/ Top Of Page China Chinese AI Chatbots Fail to Debunk Pro-China False Claims A report published by NewsGuard states that Chinese-made AI chatbots, including DeepSeek, Kimi, and Doubao, failed to debunk pro-China false claims 53% of the time, more than double the 24% failure rate of non-Chinese models tested against the same claims, with Chinese chatbots frequently responding to politically sensitive topics including Tiananmen Square, Taiwan, and Uyghur detention with evasive non-answers that leave false claims unchallenged rather than factually correcting them. As states in the report, the systematic divergence between Chinese and non-Chinese AI models on pro-China false claims indicates political alignment embedded in Chinese AI training and guardrail design, with the deflection pattern, declining to answer rather than denying, representing a technically less detectable form of bias than active repetition of false claims, since models that refuse to address sensitive topics leave users without corrections while avoiding the more easily documented behavior of directly reproducing propaganda narratives. Source: NewsGuard. As Chinese AI Models Gain Popularity in the West, Their Chatbots Fail to Debunk Pro-China False Claims More Than Half the Time. [online] Published 13 August 2026. Available at: https://www.newsguardtech.com/special-reports/chinese-ai-chatbots-fail-to-debunk-pro-china-false-claims-more-than-half-the-time Top Of Page China's PLA Deploys GoLaxy Smart Propaganda System Against Taiwan An analysis published by FPRI states that China's People's Liberation Army is deploying the GoLaxy Smart Propaganda System, an AI-powered platform that automates content generation, multi-platform distribution, and audience targeting for cognitive warfare operations against Taiwan, enabling the PLA to produce and distribute tailored influence content at a scale and speed that previously required substantially larger human operational infrastructure, as part of a broader PLA cognitive domain operations doctrine integrating information warfare with conventional military planning. According to the analysis, GoLaxy represents a structural shift in Chinese cognitive warfare capabilities against Taiwan from manual content production to AI-automated pipeline operations, enabling continuous rather than episodic influence campaigns that simultaneously target multiple audience segments, Taiwan's civilian population to undermine confidence in democratic governance and military preparedness, international Chinese-language communities to shape diaspora opinion, and English-language media to influence foreign policymaker perceptions, with automation reducing per-campaign operational cost and enabling the sustained tempo that episodic human-operated campaigns cannot maintain. Source: Foreign Policy Research Institute. Inside China’s Cognitive Warfare Playbook Against Taiwan. [online] Published 13 August 2026. Available at: https://www.fpri.org/article/2026/08/inside-chinas-cognitive-warfare-playbook-against-taiwan/ Top Of Page [AI Related Articles] Grok Produced 87% of Synthetic Audio Files A report published by Resemble AI states that its H1 2026 Deepfake Threat Report identified 15,736 confirmed victims of deepfake audio and video attacks globally in the first six months of 2026, with Grok accounting for 87% of detected synthetic audio files, a concentration attributable to xAI's comparatively permissive voice synthesis content policies relative to competing models, and with financial fraud, non-consensual intimate imagery, and political impersonation comprising the three dominant use categories. The report states that the concentration of synthetic audio production in a single platform at 87% represents a systemic vulnerability in the deepfake threat landscape, where policy decisions by one AI provider can materially alter the global distribution of harmful synthetic content, and that the 15,736 victim figure represents only confirmed cases from reports reaching Resemble AI's detection infrastructure, with researchers assessing actual victim numbers as substantially higher given the fragmented reporting environment and the volume of content that evades automated detection Source: Resemble AI. The H1 2026 Deepfake Threat Report. [online] Published 12 August 2026. Available at: https://www.resemble.ai/resources/h1-2026-deepfake-threat-report Top Of Page South Korea's KNPA Arrests people for Non-Consensual Deepfake Sexual Content An article published by SBS News states that South Korea's National Police Agency arrested 419 people in 2026 for producing or distributing non-consensual AI-generated deepfake sexual content, with those arrested primarily male students targeting female classmates using commercial AI image manipulation tools, and with the KNPA deploying a physics-based AI detection system that identifies synthetic images through lighting inconsistencies, shadow geometry errors, and physiologically impossible facial proportions invisible to human reviewers. According to the article, the KNPA's physics-based detection approach represents a technically distinct counterstrategy to existing deep-fake identification methods. Rather than training classifiers on known synthetic content patterns that degrade as generation models evolve, physics-based analysis identifies violations of physical laws that current AI generators cannot fully simulate, providing detection capability that is less vulnerable to adversarial model updates and remains effective as generation quality improves. Source: SBS News. Police Apprehend 419 Suspects Using AI to Track Deepfake Production and Creation Processes. [online] Published 13 August 2026. Available at: https://news.sbs.co.kr/english/article.do?news_id=N1008703923 Top Of Page AI Detection Tools Correctly Identified Fewer AI-Generated Election Images A report published by the Brennan Center for Justice states that AI detection tools tested on AI-generated election-related images correctly identified fewer than 4 of 14 synthetic images on average, with human reviewers performing no better than the automated tools, demonstrating that both technological and human-review safeguards for AI-generated election content are operating below reliable identification thresholds ahead of the November 2026 US midterm elections. The report states that while AI is increasingly deployed to fight election disinformation through automated monitoring and content labelling, the same AI capabilities are simultaneously enabling adversaries to produce higher volumes of more convincing disinformation at lower cost, with the detection gap documented in the report indicating that the offensive application of AI to election influence operations is currently outpacing the defensive application, creating a structural asymmetry that platform moderation systems and human reviewers are not yet equipped to close. Source: Brennan Center for Justice. Does AI Fight or Fuel Election Disinformation? [online] Available at: https://www.brennancenter.org/our-work/research-reports/does-ai-fight-or-fuel-election-disinformation Top Of Page Facebook AI-Generated Silver Fox Personas Accumulate Up to One Million Views An article published by Futurism states that Facebook is hosting networks of AI-generated fake male personas, presented as 'silver fox' relationship advisors with names including 'Uncle George,' 'Leo Davidson,' and 'The Male Insight', accumulating between 126,000 and over one million views per video, monetized through Facebook's own Content Monetization program as well as e-book sales and page resale, with Google's AI Overview initially presenting the fabricated personas as genuine people when queried, extending inauthentic identities into mainstream search results as a secondary amplification vector. According to article, the AI-generated silver fox persona networks represent a documented case of commercial-scale coordinated inauthentic behavior operating within platform monetization structures rather than against them -- with Facebook's Content Monetization program directly rewarding the engagement generated by fabricated personas, creating a financial incentive structure in which the platform's own revenue model subsidizes AI-generated inauthentic content at scale, and with Google's AI Overview propagating fabricated identities into search as a secondary credibility amplifier, illustrating how AI-generated personas can exploit both social media and search AI systems simultaneously to build false legitimacy. Source: Futurism. Facebook Is Filling With Viral AI-Generated Silver Foxes Wooing Women by Ranting About How Terrible Men Are. [online] Published 13 August 2026. Available at: https://futurism.com/artificial-intelligence/ai-generated-silver-foxes-facebook Top Of Page [General Reports] Online Anti-Migrant Narratives on X, Telegram, and TikTok Preceded Violence in Southampton and Belfast An analysis published by ISD Global states that online anti-migrant narratives circulating on X, Telegram, and TikTok directly preceded and contributed to real-world violence at asylum seeker accommodation in Southampton and Belfast in August 2026, with ISD researchers tracing the amplification pathway from initial false or misleading posts about alleged crimes by migrants to coordinated resharing by established far-right network accounts and subsequent physical violence. The analysis states that the Southampton and Belfast incidents replicate the causal amplification pathway documented during the August 2024 UK riots, in which online disinformation about a single crime allegation cascaded through coordinated far-right networks into national disorder, indicating that the enforcement actions platforms took in response to the 2024 riots have not dismantled the underlying network infrastructure, which remained fully operational and responsive to new triggering narratives in August 2026. Source: Institute for Strategic Dialogue. From Posts to Protests: Southampton and Belfast. [online] Published 20 June 2026. Available at: https://www.isdglobal.org/digital-dispatch/posts-to-protests-southampton-and-belfast/ Top Of Page Michigan Local Election Officials Anticipate Significant Disinformation Problems Ahead of November 2026 Midterms An article published by The Conversation states that a survey of Michigan local election officials found 80% confident in voting integrity ahead of the November 2026 midterm elections, while 20% anticipated significant disinformation problems, with officials citing social media-spread false claims about voter registration procedures, mail ballot rules, and polling locations as the primary threat vectors, and with smaller rural jurisdictions reporting substantially less capacity to monitor and respond to disinformation than larger county offices. As the article states, the 20% of officials anticipating significant disinformation problems disproportionately represented smaller rural jurisdictions with fewer staff resources for public communication and counter-messaging, indicating that the distribution of disinformation vulnerability among election administrators is not uniform -- with smaller offices systematically less equipped to respond to the volume and velocity of social media disinformation than major urban offices, creating pockets of elevated electoral risk in jurisdictions that national counter-disinformation resources are least likely to prioritize. Source: The Conversation. Michigan’s Local Election Officials Are Confident in Voting Integrity Ahead of November’s Midterm Election. [online] Available at: https://theconversation.com/michigans-local-election-officials-are-confident-in-voting-integrity-ahead-of-novembers-midterm-election-289133 Top Of Page Brazil Orders Discord to Suspend Livestreaming An article published by The Record states that Brazil's Supreme Court ordered Discord to suspend its livestreaming functionality after the platform failed to comply with multiple government orders to remove neo-Nazi content, channels promoting violence against minority groups, and extremist material, with the court finding Discord's content moderation response insufficient and threatening broader platform restrictions if compliance is not achieved. The article states that the Brazilian government's action against Discord represents an escalation in its regulatory approach to platform non-compliance, moving from individual content takedowns toward targeted functional suspensions that apply operational pressure while stopping short of a full platform ban, following a pattern established in Brazil's 2024 extended suspension of X for similar compliance failures, indicating that Brazilian courts are prepared to apply the same functional-suspension model to other platforms that repeatedly fail court-ordered content removal. Source: The Record. Brazil orders Discord to suspend livestreaming after teen suicide. [online] Published 13 August 2026. Available at: https://therecord.media/discord-brazil-orders-suspension-livestreaming Top Of Page Voter Registration and Biometric Verification Systems Pose Greater Structural Risk Than Deepfakes An analysis published by Rest of World states that disproportionate public focus on AI-generated deepfakes in election contexts understates the structural risk of AI systems embedded in electoral administration itself, with India's machine learning-based voter verification system linked to the Aadhaar biometric database algorithmically deleting approximately 5.5 million voters from rolls without notice across two states, a 93% algorithm error rate that required Supreme Court intervention before the program was halted and subsequently revived nationally in 2021, and with AI chatbots including ChatGPT and Gemini providing voting information to users without dedicated electoral accuracy guardrails. The analysis states that the governance gap in AI-administered elections is structurally distinct from the deepfake threat: while deepfakes operate through audience perception of fabricated content, AI embedded in voter registration, biometric verification, and results management can determine who is permitted to participate before any ballot is cast, with algorithmic exclusion operating within administrative processes whose opacity makes it systematically harder to detect, challenge, and reverse than fabricated media, and with election technologies hosted by private vendors beyond national oversight introducing accountability gaps in which no domestic legal framework governs data transfers, audits, or error correction, particularly in African electoral contexts where the AU Advisory Group has identified concentrated vendor dependency as a systemic risk. Source: Rest of World. AI Elections, Deepfakes, and Voter Chatbots. [online] Available at: https://restofworld.org/2026/ai-elections-deepfake-voter-chatbot/ Top Of Page [Appendix - Frameworks to Counter Disinformation] Nationwide Partnership Network to Fact-Check Critical 2026 Midterm Elections An article published by Poynter states that PolitiFact announced a nationwide partnership network combining state and local news organizations, university journalism programs, and digital publishers to provide coordinated fact-checking coverage of the 2026 US midterm elections, designed to extend capacity beyond PolitiFact's own reporters to smaller markets and competitive congressional districts not covered by major national fact-checkers. According to the article, the distributed partnership model directly addresses the documented geographic gap in fact-checking coverage identified in prior election cycles, in which disinformation spreading in smaller media markets and competitive House districts received substantially less verification attention than claims circulating in high-profile Senate and gubernatorial races, with local partners' constituency-specific knowledge enabling faster response to localized disinformation narratives that national fact-checkers lack the contextual expertise to evaluate quickly. Source: Poynter. PolitiFact Announces Nationwide Partnership Network to Fact-Check Critical 2026 Midterm Elections. [online] Published 13 August 2026. Available at: https://www.poynter.org/news-release/2026/politifact-announces-nationwide-partnership-network-to-fact-check-critical-2026-midterm-elections/ Top Of Page FTC Proposes Regulating Undisclosed AI Ideological Bias An article published by CyberScoop states that the Federal Trade Commission proposed applying Section 5 of the FTC Act, which prohibits unfair or deceptive acts or practices in or affecting commerce, to AI systems that embed or amplify ideological bias in their outputs without disclosing that bias to users, advancing a disclosure-focused theory of liability in which AI models that systematically skew responses in politically partisan directions while presenting themselves as neutral constitute a deceptive trade practice. The article states that the FTC's ideological bias proposal sits at the intersection of consumer protection and AI governance, with critics arguing that defining ideological bias as deceptive requires normative judgments about political balance outside the Commission's mandate, while proponents argue that undisclosed AI bias is analogous to undisclosed material product limitations, and that a disclosure obligation does not require the FTC to define correct political views, only to ensure users are informed when a model's outputs systematically favour particular political perspectives. Source: CyberScoop. The FTC Wants to Regulate AI for Ideological Bias. [online] Published 10 August 2026. Available at: https://cyberscoop.com/ftc-regulating-ai-ideological-bias/ Top Of Page [CRC Glossary] The nature and sophistication of the modern Information Environment is projected to continue to escalate in complexity. However, across academic publications, legal frameworks, policy debates, and public communications, the same concepts are often described in different ways, making collaboration, cooperation, and effective action more difficult. To ensure clarity and establish a consistent frame of reference, the CRC is maintaining a standard glossary to reduce ambiguity and promote terminological interoperability. Its scope encompasses foundational concepts, as well as emerging terms relating to Hostile Influence and Cyfluence. As a collaborative project maintained with input from the community of experts, the CRC Glossary is intended to reflect professional consensus. We encourage you to engage with this initiative and welcome contributions via the CRC website. Top Of Page

Introduction Executive Summary This installment in our series on Chinese Digital Influence in Southeast Asia examines Vietnam, a country whose relationship with China is shaped by over two millennia of conquest, cultural exchange, and independence struggles, punctuated by the 1979 Sino-Vietnamese War. Despite normalized relations since 1991, Vietnam maintains a careful balancing act: pursuing security ties with the U.S., Japan, and South Korea while upholding a policy of assurance toward China to deter coercion. China's digital influence campaign relies on state media (CGTN, Xinhua, China Radio International) promoting narratives of shared heritage and cooperation, at times spreading misleading COVID-19 origin claims. Notably, our research uncovered a coordinated inauthentic network of thousands of accounts on X, created to amplify pro-China messaging around Xi Jinping's December 2023 visit to Vietnam. China's influence also extends to diplomatic training, cyber activity, and economic leverage as Vietnam's top trading partner, ranking 43rd on the Doublethink Lab China Index for 2021.

Introduction Executive Summary In the third part of our ongoing series on Chinese digital influence in Southeast Asia our focus is on the Philippines. The aim of this paper is to create a foundation on which more detailed research can be conducted as specific case studies. With increasing tensions between China and the Philippines in the South China Sea more research is urgently needed on Chinese influence operations targeting both the Philippines and the discourse around the South China Sea conflict. Key Findings: Members of the Chinese Embassy actively use their social media accounts to spread Chinese narratives in the Philippine information space. Official accounts are amplified by inauthentic accounts. There are examples of inauthentic coordinated behavior in the Philippines. The inauthentic pro-Chinese entity Asia Today is pushing out Chinese narratives around the South China Sea conflict. Historical Background Following its independence in 1946, the Philippines established diplomatic ties with the Republic of China and maintained them even after the nationalist government relocated to Taiwan in 1949. It wasn't until 1975 that the Philippines formally recognized the People's Republic of China and normalized diplomatic relations. Over the ensuing decades, the two countries engaged in numerous high-level visits and bilateral agreements focused on trade and cultural exchange, fostering what appeared to be a growing friendship............

Introduction Executive Summary In the fourth part of our ongoing series on Chinese Influence in Southeast Asia, our Focus shifts to Malaysia. As with all papers in this series, the aim is to create a foundation upon which more detailed research can be conducted in the form of specific case studies. Numerous factors contribute to Malaysia being a focal point for influence operations. The nation boasts the fifth largest economy in Southeast Asia, representing approximately 12% of the region's GDP. With substantial deposits of tin and oil along with ranking as the 11th largest producer of bauxite and manganese, Malaysia holds significant natural resource wealth. Additionally, it also holds the third largest gas reserve in the Asia-Pacific region, following China and Indonesia. Malaysia's economic landscape is characterized by robust agricultural and manufacturing sectors. Geopolitically, Malaysia occupies a pivotal position at the intersection of the South China Sea and Indian Ocean, making it also strategically significant. The Malacca Strait, a crucial maritime route linking the two regions, traverses along the Malaysian coastline. Moreover, Malaysia faces territorial disputes, particularly with China's claims based on the controversial nine-dash line, which encroaches upon Malaysian territories in the states of....

Introduction Executive Summary In the second part of our ongoing series on Chinese digital influence in Southeast Asia we are taking a closer look at Indonesia. Indonesia has the largest population and economy in Southeast Asia and is the most populous Islamic country in the world. It boasts large deposits of natural resources such as oil, gas, nickel, gold, tin, and copper. It is one of the largest agricultural exporters in the world with nearly 30% of the land in the country used for agriculture and has a fast-expanding fisheries Industry. Under the guidance of President Jokowi Widodo Indonesia seen unprecedented economic growth an ongoing modernization of the countrys infrastructure and an increasingly flourishing tech sector, which has seen companies such as Tokopedia and Gojek arise. Geographically the country is a strategically important crossroad between the South China Sea and Indian Ocean. All major shipping lanes between the 2 regions, mainly through the Malacca Strait, need to pass along the Indonesian coastline. The demographic, economic and geographic factors explain much of the activity of Chinese influence in Indonesia. The country is a key regional player and increasingly taking up a prominent role on the international field, making it a priority for world powers such as China. Key findings: China is using both official and unofficial digital channels to push its narratives in Indonesia. Key narratives are focused on shading a positive light on the treatment of Uighurs in Xinjiang, discrediting Japan around the Fukushima incident and..........

[Introduction] Cyber-based hostile influence campaigns are aimed at influencing target audiences by promoting information and/or disinformation over the internet, sometimes combined with cyber-attacks which enhance their effect (hence force Cyfluence, as opposed to cyber-attacks that aim to steal information, extort money, etc.) Such hostile influence campaigns and operations can be considered an epistemological branch of Information Operations (IO) or Information Warfare (IW). Typically, and as customary during the last decade, the information is spread throughout various internet platforms, which are the different elements of the hostile influence campaign, and as such, connectivity and repetitiveness of content between several elements are the main core characteristics of influence campaigns. Hostile influence campaigns, much like Cyber-attacks, have also become a tool for rival nations and corporations to damage reputation or achieve various business, political or ideological goals. Much like in the cyber security arena, PR professionals and government agencies are responding to negative publicity and disinformation shared over the news and social media. We use the term cyber based hostile influence campaigns, as we include in this definition also cyber-attacks aimed at influencing (such as hack and leak during election time), while we exclude of this term other types of more traditional kinds of influence such as diplomatic, economic, military etc. During the 3rd to 9th August 2026, we observed, collected and analyzed endpoints of information related to cyber based hostile influence campaigns (including Cyfluence attacks). The following report is a summary of what we regard as the main events. Some of the mentioned campaigns have to do with social media and news outlets solemnly, while others leverage cyber-attack capabilities. [Contents] [Introduction] [Report Highlights] [Report Summary] [State Actors] Russia Storm-1516 Fabricated Video Ukraine Identifies Russian Propaganda Network Ukraine Russian SVR Distributes Second Summer 2026 Propaganda Wave China China-Linked Network Deploys Bioweapon Narrative Targeting U.S. [AI Related Articles] Five Major AI Chatbots Reproduced Content Of EU-Sanctioned Outlet EU AI Act Transparency Enforcement Begins August 2nd AI-Generated Fake Influencer Accounts More Than Half of AI-Generated Security Patches Are Broken [General Reports] Political Leaders Will Pressure Officials to Change Election Outcomes [Appendix - Frameworks to Counter Disinformation] European Commission Begins Enforcing AI Act Obligation Roadmap for Addressing Effects of Censorship, Surveillance, and New Technology on Civil Society [CRC Glossary] [ Report Highlights] Storm-1516 fabricated a video using forged Deutsche Welle and Handelsblatt branding to falsely claim German Chancellor Friedrich Merz planned to resign, attributed by German security services to Russian military intelligence, while Ukraine's Foreign Ministry documented over 1,500 pro-Russian articles exploiting Spain's Ceuta migration crisis. The Russian SVR distributed a statement claiming Ukraine has no EU prospects across 510 resources in 20 languages in its second summer 2026 propaganda wave. FDD identified a China-linked network of at least 50 coordinated assets across 10 platforms deploying hashtags including #USBiolabs and #DTRASecrets to claim U.S. biological laboratories caused COVID-19, with accounts posting identical content within hours of each other, reviving a longstanding operation to shift pandemic blame to Washington while incorporating sub-narratives targeting U.S.-Taiwan relations. The IRGC issued a statement praising Iranian revolutionary media for countering adversary "hybrid and cyber warfare including AI" while asserting that the US and Israel suffered strategic defeat in recent conflicts with Iran. Research testing five major AI chatbots across 50 propaganda narratives in three languages found 16.6% of responses engaged with content from the Foundation to Battle Injustice, an EU-sanctioned outlet founded by deceased Wagner Group leader Yevgeny Prigozhin, reproducing claims including that Ukraine reused soldier corpses for meat products, with only 52% of responses rejecting the false claims and the vulnerability attributed to generative engine optimisation techniques designed to embed propaganda into AI training data. EU AI Act transparency enforcement commenced 02-08-2026, requiring chatbots to disclose AI status at first contact, deepfakes to carry labels, and AI-generated content to include machine-readable markers; investigations were opened against OpenAI, Anthropic, Google, and X, with the European Commission publishing a list of over 180 organisations signed to the Code of Practice on transparency of AI-generated content and penalties of up to EUR 15 million or 3% of global annual turnover for non-compliance. A Gallup poll of more than 23,000 U.S. voters found 67% concerned that political leaders will pressure election officials to change election outcomes, the highest-ranked of six tested concerns, with at least 40% expressing concern about each tested aspect of election integrity and Trump's net approval rating at -20.6 points, the lowest of his second term. [ Report Summary] German security services attributed to Storm-1516, a Russia-linked operation assessed as probably connected to Russian military intelligence, a fabricated video using forged Deutsche Welle and Handelsblatt branding to claim falsely Chancellor Friedrich Merz planned to resign; both outlets confirmed the fabrication, Merz's office stated the claims were completely invented and designed to confuse and unnerve the public, while France's Viginum agency has documented Storm-1516 conducting at least 77 disinformation operations targeting Western countries between late 2023 and March 2025. Ukraine's Foreign Ministry identified a coordinated Russian propaganda network exploiting Spain's Ceuta migration crisis, with pro-Russian outlets Pravda and RT publishing over 1,500 articles including approximately 300 in Spanish targeting European audiences; Ukrainian Foreign Minister Andriy Sibiga stated Russia has for years weaponised migration and socially sensitive topics to systematically destabilise Europe and called for complete bans on Russian propaganda platforms within European jurisdictions. The Russian Foreign Intelligence Service on 03-08-2026 issued a statement claiming Ukraine has no EU membership prospects and accusing the EU of stoking the war; Ukraine's Center for Strategic Communications documented over 750 distributions across 510 resources in 20 languages targeting audiences in Russia, the USA, Europe, Africa, and Asia, the second such SVR propaganda wave of summer 2026, while the EU's diplomatic arm found 80 of 500 investigated information manipulation incidents between January 2025 and May 2026 were linked to Ukraine's EU accession process. The IRGC issued a statement praising Iranian revolutionary media for countering adversary "hybrid and cyber warfare including AI" while asserting that the US and Israel suffered strategic defeat in recent conflicts with Iran. FDD's Center on Cyber and Technology Innovation identified at least 50 coordinated assets operating across 10 platforms including TikTok, Facebook, YouTube, and X, spreading claims that U.S. biological laboratories developed bioweapons causing COVID-19, deploying hashtags including #USBiolabs, #DTRASecrets, and #AmericanBioHazard; accounts posted identical or near-identical content within hours of each other, reviving a longstanding Chinese effort to shift pandemic blame to Washington while incorporating sub-narratives targeting U.S.-Taiwan relations. Research testing five major language models, xAI's Grok, OpenAI's GPT, Anthropic's Claude, Google Gemini, and Mistral AI, across 50 propaganda narratives in three languages found 16.6% of responses engaged with content from the Foundation to Battle Injustice (r-FBI), an EU-sanctioned outlet founded by deceased Wagner Group leader Yevgeny Prigozhin; specific models reproduced claims about Ukrainian soldiers' corpses being processed for meat and Zelenskyy's cryptocurrency mining, with only 52% of responses rejecting the false claims outright. EU AI Act transparency enforcement commenced on 02-08-2026, requiring chatbots to identify themselves as automated systems, deepfakes to carry labels, and AI-generated content to include machine-readable markers; general-purpose AI model providers must publish training data summaries covering chemical, biological, radiological, and nuclear risks, with penalties of up to EUR 15 million or 3% of global annual turnover, and investigations open against OpenAI, Anthropic, Google, and X. Futurism documented AI-generated fake influencer accounts on Facebook and Instagram deploying sexually suggestive video personas to attract male users who appear unaware the figures are artificial, with individual posts drawing close to 1,000 comments; the accounts systematically exploit a post-COVID loneliness epidemic and one account's file name referenced an AI music generation tool, illustrating that years of public discourse about AI-generated media have not produced meaningful general user skepticism. 1Password researchers testing OpenAI's ChatGPT 5.5 and Anthropic's Claude Opus 4.8 on six high-impact CVEs found a 47% overall success rate for complete vulnerability patching without introducing new problems; models frequently addressed only a subset of vulnerable code paths or added fragile guard code, and Veracode's separate testing of 100 models found 44% introduced a detectable OWASP Top 10 vulnerability into the codebase, while newer frontier models with advanced cybersecurity capabilities are being distributed to key industries before open-source alternatives can compete. Researchers by the UK Ministry of Defense developed an AI system that detects online discourse manipulation by measuring how far actual replies deviate from expected conversational responses, achieving 77% accuracy, roughly twice existing methods. A Gallup poll of more than 23,000 voters conducted April 24 through June 10 found 67% concerned that political leaders will pressure election officials to change election outcomes, the top-ranked of six tested threats, followed by 57% worried about improper ballot handling, 52% about voter intimidation, and 50% that their ballots will not be counted; CFR's James M. Lindsay noted that at least 40% expressed concern about each tested aspect and that Trump's net approval rating reached -20.6 points, the lowest of his second term. The European Commission began enforcing the AI Act's first binding obligations on 2 August 2026, covering rules on prohibited AI practices (effective February 2026) and new requirements for general-purpose AI models including transparency obligations for providers of AI systems classified as presenting systemic risk; enforcement is led by the AI Office within the Commission, with national competent authorities designated across EU member states, and fines reaching up to 35 million euros or 7% of global annual turnover for the most serious violations. An interview published by Jonathon W. Penney (Osgoode Law School/Citizen Lab) argues chilling effects — self-censorship from surveillance or legal threats — stem from psychological conformity, not rational calculation, making First Amendment doctrine inadequate. He cites four amplifiers (observation, uncertainty, personalization, power asymmetry) and evidence like Wikipedia traffic drops post-Snowden. He links this to democratic withdrawal, inequality, and extremist polarization, and calls for structural reforms (antitrust, transparency, surveillance limits) before narrower fixes like anti-SLAPP laws. [State Actors] Russia Storm-1516 Fabricated Video An article published by The Local states that German security services attributed to Storm-1516, a Russia-linked operation assessed as probably connected to Russian military intelligence, a fabricated video using forged Deutsche Welle and Handelsblatt branding to claim falsely Chancellor Friedrich Merz planned to resign; both media outlets confirmed the fabrication, the CDU party dismissed it as nonsense, and Merz's office stated the claims were completely invented and designed to confuse and unnerve the public. The article states that Germany's domestic intelligence agency assessed Storm-1516 to follow political developments in Germany and exploit them opportunistically closely, with France's Viginum agency having documented the group conducting at least 77 disinformation operations targeting Western countries between late 2023 and March 2025, including prior operations during Germany's February 2025 general election that brought Merz to power, consistent with the network's documented pattern of targeting electoral and political transition periods. Source: The Local. Russia-linked group blamed for fake Merz resignation video. [online] Published 8 August 2026. Available at: https://www.thelocal.com/20260808/russia-linked-group-blamed-for-fake-merz-resignation-video Top Of Page Ukraine Identifies Russian Propaganda Network A report published by UA.NEWS states that Ukraine's Foreign Ministry identified a coordinated Russian propaganda network exploiting Spain's Ceuta migration crisis, with pro-Russian outlets Pravda and RT deploying over 1,500 articles including approximately 300 in Spanish targeting European audiences; Ukrainian Foreign Minister Andriy Sibiga characterised the strategy as exploiting topics that evoke strong emotions in society to manufacture social discord, stating the Kremlin has for years weaponised migration and similar socially sensitive topics to systematically destabilise Europe. The report states that the Ukrainian assessment characterised the campaign as a pattern of systemic destabilisation rather than an isolated incident, the network functioning as a coordinated amplification system designed to manufacture the appearance of organic European public concern over migration and called for European nations to escalate counteraction including complete bans on Russian propaganda platforms operating within European jurisdictions. Source: UA.NEWS. Ukraine has exposed a new Russian disinformation campaign in Europe. [online] Published 31 July 2026. Available at: https://ua.news/en/world/ukrayina-vikrila-novu-informatsiinu-kampaniiu-rf-u-evropi Top Of Page Ukraine Russian SVR Distributes Second Summer 2026 Propaganda Wave A report published by UNN.ua states that the Russian Foreign Intelligence Service on 03-08-2026 issued a statement claiming Ukraine has no prospects for EU membership and accusing the EU of stoking the war, with Ukraine's Center for Strategic Communications monitoring over 750 distributions across 510 resources in 20 languages targeting audiences in Russia, the USA, Europe, Africa, and Asia, the second such SVR propaganda wave of summer 2026, following a prior statement distributed on June 19th. The report states that between January 2025 and May 2026, 80 of 500 foreign information manipulation incidents investigated by the EU's diplomatic arm were linked to Ukraine's EU accession, with the EU identifying steady patterns of narrative alignment between Russian state channels and assets that launder, repurpose, and amplify Kremlin talking points, including narratives that the EU is prolonging the war to weaken Russia, that some EU countries want to partition Ukraine, and that European and Ukrainian values are fundamentally incompatible. Source: UNN. Russia has launched a new wave of propaganda against Ukraine’s accession to the EU. [online] Available at: https://unn.ua/en/news/russia-has-launched-a-new-wave-of-propaganda-against-ukraine-s-accession-to-the-eu Top Of Page China China-Linked Network Deploys Bioweapon Narrative Targeting U.S. An analysis published by FDD states that FDD's Center on Cyber and Technology Innovation identified at least 50 coordinated assets across 10 platforms, including TikTok, Tumblr, Facebook, YouTube, Blogspot, X, Quora, and Pixiv, spreading claims that U.S. biological laboratories developed bioweapons responsible for COVID-19, deploying hashtags including #USBiolabs, #DTRASecrets, and #AmericanBioHazard while alleging the U.S. Defense Threat Reduction Agency manages a covert military laboratory network for infectious disease development; accounts repeatedly posted identical or near-identical text, hashtags, cartoons, videos, and images within hours of each other. The analysis states that the network revives a longstanding Chinese effort to shift pandemic blame to Washington, incorporating sub-narratives tailored to divide Washington and Taipei, including a 2023 Chinese state media amplification of a forged document falsely claiming the U.S. directed Taiwan to build a biological-weapons laboratory, while other posts alleged U.S.-supported facilities stored high-risk pathogens and destroyed samples in Ukraine to conceal evidence, illustrating a multi-theatre narrative architecture deploying distinct regional sub-narratives from a single coordinated asset base. Source: Foundation for Defense of Democracies. China-Linked Social Media Network Seeks To Shift COVID Responsibility From China to U.S. [online] Published 7 August 2026. Available at: https://www.fdd.org/analysis/2026/08/07/china-linked-social-media-network-seeks-to-shift-covid-responsibility-from-china-to-u-s/ Top Of Page Iran IRGC Praises Iranian State Media as Counter-Disinformation Frontline An article published by Tasnim News states that the Islamic Revolution Guards Corps stated on August 8th, 2026, praising Iranian revolutionary media outlets for countering what the IRGC characterized as adversary deployment of "hybrid and cyber warfare capabilities, including artificial intelligence" to distort facts and create division between the Iranian public and the Islamic Republic, framing Iranian state media as a frontline instrument in an active information conflict against the United States and Israel. The article states that the IRGC statement claimed the United States and Israel failed to achieve their objectives in recent conflicts with Iran, that international media acknowledged US "failure" and described Trump's actions as a "defeat", and that post-conflict understandings amounted to an "American surrender document", claims presented without independent verification and consistent with IRGC information operations framing; the statement is notable as a primary source document revealing how Iran officially positions its state media apparatus as a counter-disinformation instrument while simultaneously using that apparatus to propagate unverified narratives about adversary strategic failure. Source: Tasnim News Agency. IRGC Lauds Iranian Media for Countering Enemy Disinformation. [online] Published 8 August 2026. Available at: https://www.tasnimnews.ir/en/news/2026/08/08/3665982/irgc-lauds-iranian-media-for-countering-enemy-disinformation Top Of Page [AI Related Articles] Five Major AI Chatbots Reproduced Content Of EU-Sanctioned Outlet An investigation published by Euronews states that researchers testing five major language models, xAI's Grok, OpenAI's GPT, Anthropic's Claude, Google Gemini, and Mistral AI, across 50 propaganda narratives in three languages (English, German, French) found 16.6% of responses engaged with material from the Foundation to Battle Injustice (r-FBI), an EU-sanctioned outlet founded in 2021 by deceased Wagner Group leader Yevgeny Prigozhin that presents itself as a human rights NGO; 30.9% of responses addressed the topics without disclosing the sanctioned source, while only 52% rejected or debunked the false claims. The investigation states that the underlying vulnerability involves generative engine optimization, techniques designed to influence AI training data, paralleling search engine optimization methods, with specific examples including OpenAI's GPT 4.1 Mini citing r-FBI articles supporting claims about Ukrainian soldiers' corpses being processed for meat products, and Mistral Small 3.2 endorsing allegations about President Zelenskyy's cryptocurrency ventures causing Ukraine's energy crisis, identifying AI training data infiltration as an emerging hostile influence vector distinct from direct content injection. Source: Euronews. AI chatbots citing Russian propaganda sourced from EU-sanctioned outlet. [online] Published 27 July 2026. Available at: https://www.euronews.com/my-europe/2026/07/27/ai-chatbots-citing-russian-propaganda-sourced-from-eu-sanctioned-outlet Top Of Page EU AI Act Transparency Enforcement Begins August 2nd An analysis published by Help Net Security states that EU AI Act transparency enforcement commenced on August 2nd, 2026, requiring chatbots to identify themselves as automated systems at first contact, deepfakes to carry labels, and AI-generated or altered content to include machine-readable markers; general-purpose AI model providers must document information for competent authorities, implement copyright policies, and publish training data summaries addressing risks including chemical, biological, radiological and nuclear incidents, loss of control, cyber offences, harmful manipulation, and fundamental rights threats, with penalties of up to EUR 15 million or 3% of worldwide annual turnover for non-compliance. An analysis published by Help Net Security states that enforcement is being pursued against OpenAI, Anthropic, Google, and X among others, with over 180 organisations having signed the Code of Practice on transparency of AI-generated content; subsequent enforcement milestones include a December 2nd, 2026 ban on non-consensual AI-generated sexually explicit content and child sexual abuse material, a December 2nd, 2027 deadline for high-risk AI systems, and a August 2nd, 2028 deadline for high-risk systems built into regulated products, marking a phased escalation from transparency requirements to systemic risk regulation. Source: Help Net Security. EU Begins Enforcing AI Act, Putting AI Models Under the Microscope. [online] Published 4 August 2026. Available at: https://www.helpnetsecurity.com/2026/08/04/eu-ai-act-enforcement-ai-models/ Top Of Page AI-Generated Fake Influencer Accounts A report published by Futurism states that Meta's platforms are filling with AI-generated video of nonexistent young women deployed by fake influencer accounts, including accounts called Grace the Gymnast, Kylie Blaze, and Issy Dexan, that use sexually suggestive language to lure in male users who appear unaware the personas are artificial, with individual posts drawing close to 1,000 comments; the accounts systematically exploit a post-COVID loneliness epidemic, with experts finding that social media use makes users lonelier rather than fostering human connection. A report published by Futurism states that the AI-generated fake influencer schemes represent a convergence of easy-to-use video generation tools with social engineering techniques targeting basic psychological instincts, with the operations growing more sophisticated since 2023, moving from still images to AI video, and the widespread unaware engagement illustrating that years of public discourse about AI-generated media have not produced the population-level skepticism that media literacy initiatives predicted. Source: Futurism. Boomers on Facebook Are Leaving Comments That Make It Clear That They Have No Idea That AI-Generated Thirst Traps Aren’t Real. [online] Published 4 August 2026. Available at: https://futurism.com/artificial-intelligence/boomers-facebook-ai-thirst-traps Top Of Page More Than Half of AI-Generated Security Patches Are Broken An article published by CyberScoop states that 1Password researchers testing OpenAI's ChatGPT 5.5 and Anthropic's Claude Opus 4.8 on six high-impact, high-complexity CVEs found an overall success rate of 47% for complete vulnerability patching without introducing new problems, below a coin flip, with models frequently addressing only a subset of vulnerable code paths, adding fragile guard code that satisfied tests while failing to address the vulnerability's root cause, and sometimes introducing subtle changes in application behaviour while patching the immediate issue. The article states that Veracode's separate testing of 100 models found 44% introduced a detectable OWASP Top 10 vulnerability into the codebase, with most models showing stagnation in security patching despite enormous strides in crafting workable code; a Veracode vice president noted that AI tools remain unreliable for cybersecurity patching without knowledgeable humans in the loop, and that some kind of automated code review will be necessary, preferably not by the same automation tool that produced the code, in a world where AI agents generate exponentially more code than human defenders can review. Source: CyberScoop. More than half of AI-generated patches are broken. [online] Published 7 August 2026. Available at: https://cyberscoop.com/ai-code-patching-security-risks/ Top Of Page Researchers Develop AI Detection System That Identifies Discourse Derailment in Online Comments An article published by The Conversation states that researchers at Cardiff University and the University of Sheffield, funded by the UK Ministry of Defense Science and Technology Laboratory, developed a detection system for AI-assisted discourse derailment in online comment sections, finding that 65% of identified manipulative comments contained non sequiturs, 36% red herrings, and 20% personal attacks, with derailing comments significantly less likely to acknowledge prior messages or demonstrate empathy, based on manual analysis of over 1,600 comments beneath BBC News YouTube videos labeled against 25 discussion features. The article states that the detection system works by measuring divergence between actual replies and a set of AI-generated expected responses, with substantial deviation flagging potential discourse manipulation rather than scanning for suspicious vocabulary; the system achieved approximately 77% accuracy, roughly twice the effectiveness of existing word-level sentiment analysis systems and matching human researcher agreement levels, with the authors noting that as AI lowers the cost and raises the volume of synthetic disinformation, detection methods capable of identifying manipulation through conversational pattern rather than content alone become operationally significant. Source: The Conversation. AI is making disinformation harder to spot, but we’ve found a new way to catch it. [online] Available at: https://theconversation.com/ai-is-making-disinformation-harder-to-spot-but-weve-found-a-new-way-to-catch-it-288671 Top Of Page [General Reports] Political Leaders Will Pressure Officials to Change Election Outcomes An analysis published by CFR states that a Gallup poll of more than 23,000 voters conducted from April 24th, 2026 through June 10th, 2026 found 67% of Americans concerned that political leaders will pressure election officials to change election outcomes, the top-ranked threat, followed by 57% worried about improper ballot handling, 52% worried about voter intimidation near polling places, and 50% worried their ballots will not be counted; at least 40% of the public expressed concern about each of the six tested aspects of elections, with a majority concerned about four of them. According to the analysis, Trump's net approval rating now stands at -20.6 points, the lowest of his second term, while Democrats hold a 6.4-point edge on the generic congressional ballot; Larry Sabato's Crystal Ball projects 20 House seats as toss-ups, 15 currently or formerly Republican-held, up from April 13th as seats shift from leans Republican to toss-up, with CFR's Lindsay noting that democracy cannot succeed if the governed doubt that election results reflect their will. Source: Council on Foreign Relations. Voters Worry About Election Integrity. [online] Published 3 August 2026. Available at: https://www.cfr.org/articles/voters-worry-about-election-integrity Top Of Page [Appendix - Frameworks to Counter Disinformation] European Commission Begins Enforcing AI Act Obligation An announcement published by the European Commission states that the AI Act's first binding obligations began enforcement on August 2nd, 2026, covering rules on prohibited AI practices that took effect in February 2026 and new requirements for general-purpose AI models including transparency obligations for providers of systems classified as presenting systemic risk; enforcement is led by the AI Office within the Commission, with national competent authorities designated across EU member states and fines reaching up to 35 million euros or 7% of global annual turnover for the most serious violations. The announcement states that general-purpose AI model providers must now document their training data, energy consumption, and safety evaluations and submit these to the AI Office; providers of models designated as systemic-risk AI, defined as models trained with more than 10^25 floating-point operations, face additional adversarial testing, incident reporting, and cybersecurity requirements, with the Commission indicating it has already initiated formal evaluation processes against several major providers. Source: European Commission. Commission starts enforcing AI Act rules and new transparency requirements on 2 August. [online] Published 31 July 2026. Available at: https://digital-strategy.ec.europa.eu/en/news/commission-starts-enforcing-ai-act-rules-and-new-transparency-requirements-2-august Top Of Page Roadmap for Addressing Effects of Censorship, Surveillance, and New Technology on Civil Society An interview published by TechPolicy.Press states that Jonathon W. Penney, York Research Chair in AI, Data Governance and the Law at Osgoode Law School and research fellow at Citizen Lab argues that chilling effects, defined as self-censorship and behavioral restraint triggered by awareness of surveillance, legal threats, or personal targeting, operate through deep psychological conformity mechanisms rather than rational legal calculation, rendering conventional First Amendment doctrine structurally inadequate; Penney identifies four amplifying factors, observation, uncertainty, personalization, and power asymmetry and supports the framework with population-scale empirical data including Wikipedia traffic declines following the 2013 Snowden disclosures and Google search behavior shifts for sensitive terms. The interview states that Penney identifies secondary harms flowing from chilling effects including democratic withdrawal, entrenchment of inequality through disproportionate surveillance of marginalized populations, and polarization within extremist subgroups where conformity mechanisms amplify rather than suppress hostile behavior; his policy framework prioritizes structural reform, antitrust enforcement against tech consolidation, government transparency, and surveillance restriction as preconditions before piecemeal measures such as expanded anti-SLAPP statutes, disclosure requirements, and personalized-targeting restrictions can be effective, framing the underlying problem as one of power: "privacy is about power... chilling effects is also a struggle over power, information, and broader structural power relationships in society". Source: Tech Policy Press. A Roadmap for Confronting the Chilling Effects of Censorship, Surveillance and New Technology. [online] Published 26 July 2026. Available at: https://www.techpolicy.press/a-roadmap-for-confronting-the-chilling-effects-of-censorship-surveillance-and-new-technology/ Top Of Page [CRC Glossary] The nature and sophistication of the modern Information Environment is projected to continue to escalate in complexity. However, across academic publications, legal frameworks, policy debates, and public communications, the same concepts are often described in different ways, making collaboration, cooperation, and effective action more difficult. To ensure clarity and establish a consistent frame of reference, the CRC is maintaining a standard glossary to reduce ambiguity and promote terminological interoperability. Its scope encompasses foundational concepts, as well as emerging terms relating to Hostile Influence and Cyfluence. As a collaborative project maintained with input from the community of experts, the CRC Glossary is intended to reflect professional consensus. We encourage you to engage with this initiative and welcome contributions via the CRC website. Top Of Page