top of page
CRC logo (2).png

Cyber based influence campaigns 17th - 23rd August 2026 Report

  • Writer: CRC
    CRC
  • 15 minutes ago
  • 26 min read
Cover Image- Text: Weekly Media Update: Information Operations


[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 17th to 23rd 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]





Tiktok 


Russia 

China







[ Report Highlights]


  • Russia's Operation Matryoshka and Storm-1516 simultaneously targeted Germany, France, and Sweden in August 2026 using fake broadcaster logos, with France opening the first criminal prosecution directly targeting named political candidates (Attal, Philippe, Glucksmann), and the Swedish operation releasing seven fabricated clips within 60 minutes designed to overwhelm newsroom fact-checking capacity rather than maximize mass audience reach.

  • The PRC deployed an AI-enabled autonomous cyberattack against 21 Taiwanese government systems in July 2026 while simultaneously running a disinformation campaign targeting Kaohsiung's 2026 mayoral election, with Taiwan's DPP warning that AI has lowered content fabrication costs to the point where Beijing can generate false narratives faster than Taiwanese fact-checkers can respond, describing PRC interference as a "comprehensive election interference industry supply chain".

  • Russia is running a coordinated foreign military recruitment pipeline through paid Meta advertisements and 35 or more websites in 40 languages targeting Africa, Latin America, and Southeast Asia with misleading military job offers, with 85 paid Meta ads flagged during the reporting period.

  • AI-generated fake news websites have proliferated to the point where they are indistinguishable from authentic outlets to most readers, while US Special Operations Command warned publicly that AI-enabled deception has made the information environment so opaque that senior officials can no longer reliably distinguish authentic from fabricated content, directly raising the probability of large-scale military conflict.

  • A fake polling company (Median Strategies) admitted its surveys published across at least three US states were a 'social experiment', demonstrating a structural vulnerability; polling data published under a professional-sounding name receives less source verification than individual factual claims because polling is treated as systematic measurement, creating a documented entry point for fabricated quantitative data into the political information environment.

  • The EU activated its DSA crisis coordination mechanism with Meta and TikTok following the Ceuta border crisis, in which AI-generated videos falsely claiming the border was open contributed to approximately 100 deaths, marking the first operational use of the EU's voluntary crisis fact-checking framework, whose activation depends on platform voluntary participation rather than legal obligation.

  • Detector Media documented a coordinated August 12th – 18th fabrication campaign combining a manipulated Polish criminality compilation, a fake BBC video using AI-generated narration over genuine Elysee Palace footage to fabricate a Zelenska jewelry scandal, and an entirely neural-network-generated assault photograph, while Ukrinform separately debunked a fabricated audio clip falsely claiming Dnipro banned evacuation of conscription-age men.

  • The FTC's Operation AI Comply has expanded to business-to-business enforcement using personal liability doctrine two years after its launch, with the SEC launching a parallel AI washing enforcement campaign, marking a structural shift from consumer protection to market integrity framing of AI misrepresentation, with corporate decision-makers now facing personal liability under the means and instrumentalities doctrine.

[ Report Summary]

  • The EU activated its voluntary DSA crisis fact-checking coordination mechanism with Meta and TikTok following the Ceuta border crisis, in which AI-generated videos falsely claiming the border between Morocco and the Spanish enclave of Ceuta was open contributed to a mass crossing that resulted in approximately 100 deaths. Spanish fact-checker Newtral and Moroccan fact-checker MapExpress were coordinated by the EU to flag and escalate false content to the platforms during the crisis.

  • EUvsDisinfo documented a coordinated Russian foreign military recruitment operation using paid Meta advertisements, WhatsApp messaging, and at least 35 websites in up to 40 languages to recruit foreign fighters with misleading job offers. African audiences were targeted from June 2026, Latin American audiences and European diaspora communities from July, and Southeast Asian audiences from late July. The investigation flagged 85 paid Meta ads during the reporting period.

  • French Foreign Minister Jean-Noël Barrot published a signed opinion piece in Le Monde arguing that the expulsion of Russian national Xenia Fedorova, whose residence permit was revoked on 29th July 2026 after French authorities determined she had been executing Kremlin-directed destabilisation operations through appearances on CNews and Europe 1 and published columns, illustrates the scale of the foreign interference threat France currently faces, describing Fedorova as a 'professional of subversion'. Fedorova responded in Le JDD, challenging Barrot to produce evidence of the Russian interference attributed to her.

  • French prosecutors opened a criminal investigation into a Russian-linked disinformation campaign fabricating health claims against three centrist presidential candidates: a fabricated Parkinson's disease claim against former Prime Minister Gabriel Attal, a health rumour against former PM Edouard Philippe, and false information about MEP Raphael Glucksmann. France's digital interference watchdog Viginum attributed the Attal campaign to Operation Matryoshka and the Philippe and Glucksmann operations to Storm-1516.

  • A Russia-linked influence network deployed seven fabricated video clips within 60 minutes on 18th August 2026, depicting Swedish Prime Minister Ulf Kristersson making statements he never made and using forged visual identities of broadcaster SVT and newspaper Dagens Nyheter. Euromaidan Press attributes the operation to the same network that struck Germany with fake BBC and ARD websites targeting the September 2026 federal election and Storm-1516 campaigns targeting French candidates in August, noting that simultaneous multi-clip deployment is designed to overwhelm newsroom fact-checking capacity.

  • Russian propaganda spread a fabricated audio clip replacing the original soundtrack of a genuine 7th August 2026 TikTok video showing train delays, falsely claiming Dnipro city authorities banned the evacuation of men of conscription age. The fake first appeared 17th August 2026 via the pro-Russian Telegram channel Perimeter ZOV on X. Ukrainian law contains no provision prohibiting conscription-age men from evacuating.

  • Detector Media's review of Russian disinformation from 12th to 18th August 2026 documented three coordinated fabrication tracks: a manipulated compilation falsely portraying systematic Ukrainian criminality in Poland (Polish police data shows Ukrainians committed approximately 2% of offenses 2022-2025); a fake BBC video using AI-generated narration over genuine Elysee Palace footage to fabricate a Zelenska jewelry theft; and a fully neural-network-generated assault photograph presented as a Zelenska security incident.

  • Jamestown Foundation analyst Eitvydas Bajarunas argues that Lithuania confronts sustained Russian hybrid pressure across informational, cyber, economic, and kinetic domains simultaneously, with the primary vulnerability being institutional misalignment; Lithuanian decision-making follows peacetime logic while Russia operates under wartime assumptions. Lithuania allocates 5.38% of GDP to defense and hosts a German brigade, yet structural gaps in decision-making speed, counter-drone coverage, and total societal mobilization undermine those capabilities.

  • EUvsDisinfo documents how Kremlin-aligned outlets have largely abandoned climate change denial in favour of weaponising climate narratives against EU policy, recasting EU climate and energy regulations as authoritarian control mechanisms, predicting European economic collapse without Russian gas, and portraying Russia as a future climate beneficiary with Siberia as 'the land of the future', while Russia's own fuel crisis, driven by Ukrainian strikes on oil refineries, deepens the contradiction at the centre of its energy narratives.

  • AEI/ISW's China-Taiwan Update documented PRC targeting of Taiwan's 2026 Kaohsiung mayoral election with fabricated polling data and fake news, a July 2026 AI-enabled autonomous cyberattack that compromised 21 Taiwanese government systems exfiltrating approximately 2,500 records, and Philippine President Marcos Jr. publicly rejecting use of Philippine territory as a US launch platform in a Taiwan contingency, a position the analysis links to PRC political influence on Manila.

  • Taiwan's Democratic Progressive Party China Affairs Department warned that AI has significantly lowered the cost of producing disinformation, enabling the PRC to operate a comprehensive election interference industry supply chain that combines voter data profiling, content impersonation mimicking local vernacular, and speed-based saturation generating false content within hours before elections to outpace Taiwan's fact-checking capacity, describing the operation as a continuous infrastructure refined across multiple election cycles since 2016.

  • NewsGuard Reality Check documented the proliferation of AI-generated news websites designed to pass as authentic local and national outlets, finding that these sites are now indistinguishable from genuine news organizations to most readers, including those with media literacy.

  • Admiral Frank Bradley, Commander of US Special Operations Command, warned publicly that AI-enabled deception has made the information environment so opaque that even senior military and intelligence officials cannot reliably distinguish authentic from fabricated content, that counter-influence structures have been systematically dismantled, and that hostile active measures have 'eaten the world', directly raising the probability of large-scale military conflict by increasing the likelihood that decision-makers will act on fabricated intelligence.

  • Carnegie Endowment researchers Danaé Metaxa and Alex Engler document that LLMs have achieved political persuasiveness comparable to human-written content across 19 models and over 700 political issues, while asserting false claims as true 35% of the time on misinformation queries.

  • A forthcoming academic book by Manhattan University psychologist Jay Friedenberg, uploaded as a preprint to PhilPapers, provides a comprehensive typology of disinformation operations used by authoritarian regimes, covering propaganda mechanics, cognitive bias exploitation, social media amplification, deepfakes, and AI-generated content risks across eight chapters.

  • A peer-reviewed study based on 78 in-depth interviews with influence operation insiders in Indonesia, the Philippines, and Thailand identifies five simultaneous IO transformations: diversification of clients beyond the state to political parties and businesses; increased outsourcing to private PR firms; a shift from high-volume bot-like accounts to carefully crafted pseudonymous influencer accounts with large genuine followings; content strategies moving from repetitive state-mirroring messaging to data-analytics-driven 'gray' campaigns that reframe real news rather than fabricate disinformation; and dissemination shifting from mass flooding to responsive real-time data-informed communication.

  • NewsGuard investigation found the four largest demand-side advertising platforms, Google's Display & Video 360, Amazon DSP, Yahoo DSP, and Adobe Advertising DSP, placing programmatic advertisements for their own parent company brands and client brands next to articles advancing provably false claims.

  • A polling company named Median Strategies admitted that surveys it published across at least three US states were fabricated as a social experiment. The company had no verifiable address, staff, or methodology disclosures, and the fabricated data circulated in political media without verification before the admission, illustrating that synthetic polling data can enter the political information ecosystem through normal reporting channels because survey data receives less source verification than individual factual claims.

  • 19 US lawmakers led by Representative Kevin Mullin sent a letter to Meta CEO Mark Zuckerberg demanding specific details of the company's election safeguards against AI-generated deepfakes ahead of the 2026 midterm elections, citing documented rollbacks of Meta's previous election integrity commitments and the absence of publicly disclosed standards for detecting or removing AI-generated political impersonation content from Facebook and Instagram.

  • TechPolicy.Press documented the convergence of three developments ahead of the 2026 US midterms: AI chatbots becoming a primary electoral information source for voters despite a documented 50% factual error rate on election-related queries; political campaigns deploying synthetic voter focus groups, AI-simulated voter segment representations, to craft messaging; and official CISA guidance links to election information resources breaking following structural changes to agency communications, leaving voters seeking authoritative federal information directed to dead links.

  • Holland & Knight analyzed the evolution of the FTC's Operation AI Comply enforcement campaign two years after its September 2024 launch, documenting its expansion from B2C consumer protection to business-to-business AI capability misrepresentation using the means and instrumentalities doctrine to impose personal liability on corporate decision-makers, while the SEC has launched a parallel AI washing enforcement campaign targeting publicly traded companies that misrepresent AI capabilities to investors.

[State Actors]


Tiktok

EU Activates DSA Crisis Mechanism with Meta and TikTok After Ceuta Deaths

An article published by Euronews states that the European Union activated a voluntary crisis fact-checking coordination mechanism with Meta and TikTok under the Digital Services Act following the Ceuta border crisis, in which AI-generated videos falsely claiming the border crossing between Morocco and the Spanish enclave of Ceuta was open contributed to a mass crossing that resulted in approximately 100 deaths, with the EU working with Spanish fact-checker Newtral and Moroccan fact-checker MapExpress to flag false content to the platforms rapidly. Meta separately removed coordinated content promoting human smuggling routes.


The article states that the DSA crisis coordination mechanism activated during the Ceuta crisis represents the first operational use of the EU's voluntary Code of Practice for crisis-scale information events, establishing a precedent in which platform self-regulatory commitments are activated as an emergency response to disinformation causing direct physical harm, while simultaneously exposing the mechanism's structural limitation: activation depended on platform voluntary participation rather than legal obligation, and the approximately 100 deaths occurred before takedown requests reached sufficient scale to slow the migration.


Source: Euronews. EU Coordinates Fact-Checkers with Meta and TikTok to Avoid Another Online-Fuelled Ceuta Crisis. [online] Published 11 August 2026. Available at: https://www.euronews.com/my-europe/2026/08/11/eu-coordinates-fact-checkers-with-meta-and-tiktok-to-avoid-another-online-fuelled-ceuta-crisis 


Russia

Russia Recruits Foreign Fighters via Paid Meta Ads

An investigation published by EUvsDisinfo states that Russia's military recruitment operation targets foreign nationals through a coordinated digital pipeline comprising paid Meta advertisements, WhatsApp messaging, and at least 35 recruitment websites operating in up to 40 languages, including Indonesian, Thai, Malay, Arabic, Wolof, French, and Spanish, with African audiences targeted from June 2026, Latin American audiences and European diaspora communities targeted from July, and Southeast Asian audiences from late July, using misleading job offers and hidden military contracts that routinely changed after recruits arrived.


The investigation states that Russia's foreign recruitment pipeline represents a coordinated inauthentic behavior operation running on commercial advertising infrastructure: the 85 paid Meta ads flagged during the reporting period demonstrate that Russia is exploiting the same paid-distribution channels available to any advertiser to reach audiences facing poverty and limited employment prospects with military recruitment content designed to appear as legitimate employment offers, with the operational geography, Africa, Latin America, and Southeast Asia, selected precisely because these regions are least likely to be covered by NATO-aligned counter-influence monitoring systems.


Source: EUvsDisinfo. From Social Media to the Front Line: Russia’s Foreign Recruitment Pipeline. [online] Published 14 August 2026. Available at: https://euvsdisinfo.eu/from-social-media-to-the-front-line-russias-foreign-recruitment-pipeline/ 


Fedorova Expulsion Illustrates Scale of Russian Interference Against France

An opinion published by Le Monde states that French Foreign Minister Jean-Noël Barrot, writing in a signed ministerial opinion piece, argued that the expulsion of Russian national Xenia Fedorova, whose residence permit was revoked on 29th July 2026 following a French government determination that she had been executing Kremlin-directed destabilisation operations through her appearances on CNews and Europe 1 and her published columns, illustrates the scale of the foreign interference threat France faces, describing her as a 'professional of subversion' and an agent executing operations piloted by the Kremlin against French society, with her bank accounts frozen and financial transfers to her prohibited following the expulsion order.


Barrot's choice to frame the Fedorova expulsion through a signed ministerial op-ed in France's newspaper of record, rather than a routine administrative statement, reflects a deliberate public attribution strategy: by describing a named private individual as a Kremlin agent in a ministerial capacity, the French government is establishing a political narrative of documented Russian interference that functions independently of any judicial determination, a posture underscored by Fedorova's counter-challenge in Le JDD demanding Barrot produce evidentiary support for the interference characterisation, and by a separate France 24 report that Le Figaro's identity had been usurped in a fabricated citation falsely attributing statements about Fedorova to Barrot, suggesting Russian disinformation operations responded to the expulsion in kind.


Source: Le Monde. France's Foreign Minister: 'The Case of Xenia Fedorova Highlights the Interference Threat We Face'. [online] Published 17 August 2026. Available at: https://www.lemonde.fr/en/opinion/article/2026/08/17/france-s-foreign-minister-the-case-of-xenia-fedorova-highlights-the-interference-threat-we-face_6756598_23.html


France Opens Criminal Investigation Into Russian Disinfo Campaign Against Presidential Candidates

An article published by The Local states that French prosecutors opened a criminal investigation into a Russian-linked disinformation campaign fabricating health claims against three prominent centrist political figures, former Prime Minister Gabriel Attal (fabricated Parkinson's disease claim), former Prime Minister Edouard Philippe (fabricated health rumour), and MEP Raphael Glucksmann (false personal information), with France's national digital interference watchdog Viginum attributing the Attal campaign to Operation Matryoshka and the Glucksmann and Philippe operations to the separate Storm-1516 network, marking the first time French prosecutors have opened criminal charges against a Russian state-linked influence operation targeting named candidates.


The French criminal investigation marks a structural shift in democratic governments' response to foreign interference: rather than diplomatic protests or sanctions, France is using domestic criminal law to target the specific actors behind named disinformation operations, establishing legal precedent that fabricating health claims about identifiable political figures for electoral interference purposes constitutes a criminal act under French law, and creating an accountability mechanism that could deter future IO operations that have historically relied on the impunity that informal attribution provided.


Source: The Local France. France Probes Russian Disinfo Campaign Against Centrist Candidates. [online] Published 18 August 2026. Available at: https://www.thelocal.com/20260818/france-probes-russian-disinfo-campaign-against-centrist-candidates/


Russia Floods Swedish Media with Fake Clips to Target PM Kristersson

A report published by Euromaidan Press states that a Russian-linked influence network published seven fabricated video clips within 60 minutes on 18 August 2026, depicting Swedish Prime Minister Ulf Kristersson making statements he never made and using forged visual identities of Sweden's public broadcaster SVT and newspaper Dagens Nyheter, with the operation's rapid multi-platform deployment confirming attribution to the same network that struck Germany with fake BBC and ARD websites targeting the September 2026 federal election, and Storm-1516 fabrications targeting French candidates earlier in August.


The Swedish operation's defining tactical innovation is its objective of overwhelming newsroom capacity rather than maximising mass-audience reach: by releasing seven distinct fabricated clips within a single hour, the operation forces journalists across multiple outlets to simultaneously investigate and debunk content, consuming verification bandwidth and delaying corrections, a strategy calibrated to the reality that professional fact-checkers, not general audiences, are the primary barrier between disinformation and credible propagation through secondary reporting.


Source: Euromaidan Press. Russia Faked Broadcasts from Sweden’s National TV to Smear Its PM—the Same Network Already Hit Germany and France This Month. [online] Published 20 August 2026. Available at: https://euromaidanpress.com/2026/08/20/russia-faked-broadcasts-from-swedens-national-tv-to-smear-its-pm-the-same-network-already-hit-germany-and-france-this-month/


Russian Propaganda Fabricates Dnipro Evacuation Ban for Conscription-Age Men

A factcheck published by Ukrinform states that Russian propaganda spread a fabricated audio clip falsely claiming that Dnipro city authorities banned the evacuation of men of conscription age, with the audio replacing the original soundtrack of a genuine TikTok video from 07 August 2026 showing train delays caused by military transport, the fake first appearing on 17 August 2026 via the pro-Russian Telegram channel Perimeter ZOV on X, and Ukrainian law containing no provision prohibiting men of conscription age from evacuating.


A factcheck published by Ukrinform states that the fabricated Dnipro evacuation ban narrative is designed to amplify fear among Ukrainian civilian men that remaining in or returning to Ukraine will result in immediate mobilization, serving the dual purpose of discouraging civilian population movement and undermining trust in Ukrainian civil administration by falsely attributing draconian restrictions to local authorities, a disinformation pattern that converts genuine public concern about mobilization into a fear-based deterrent through audio manipulation of authentic citizen-generated content.


Source: Ukrinform. Russian Propaganda Spreads Fake Claim About Ban on Evacuation of Conscription-Age People in Dnipro. [online] Available at: https://www.ukrinform.net/rubric-factcheck/4155918-russian-propaganda-spreads-fake-claim-about-ban-on-evacuation-of-conscriptionage-people-in-dnipro.html


Russian Fakes Target Zelenska and Ukrainian Diaspora in Poland

A report published by Detector Media states that Russian disinformation during 12th to 18th August 2026 included a fabricated compilation of unrelated criminal incidents from multiple years designed to portray systematic Ukrainian criminality in Poland (when Polish police data shows Ukrainians committed approximately 2% of offenses in 2022-2025), a fake BBC video using AI-generated narration combined with genuine Elysee Palace footage to falsely claim First Lady Olena Zelenska wore an 800,000 euro pendant stolen from a French museum, and a fully AI-generated photograph presented as evidence of a restaurant assault by Zelenska's security staff, with SynthID markers confirming artificial generation of the taxi sign image and facial analysis confirming the assault photo was entirely neural-network generated.


The three fabrication tracks target three distinct Ukrainian credibility vulnerabilities simultaneously, the Polish criminality narrative exploits host-country anxieties about Ukrainian refugees in Europe's largest Ukrainian diaspora destination to damage bilateral relations; the Zelenska pendant fabrication attacks the perception of Ukraine's wartime leadership class as corrupt beneficiaries of Western support; and the restaurant assault narrative deploys a Prigozhin-linked character attack sustained across multiple weeks, indicating coordinated deployment calibrated to maintain saturation rather than rely on one-time exposure.


Source: Detector Media. “Crimes and Acts of Sabotage by Ukrainians” in Poland and Zelenska in a “Stolen Pendant.” Review of Russian Fakes from August 12–18, 2026. [online] Published 21 August 2026. Available at: https://en.detector.media/post/crimes-and-acts-of-sabotage-by-ukrainians-in-poland-and-zelenska-in-a-stolen-pendant-review-of-russian-fakes-from-august-12-18-2026


Lithuania's Institutional Lag Creates Exploitable Vulnerabilities Under Russian Hybrid Pressure

An analysis published by Jamestown Foundation states that Lithuania confronts sustained Russian hybrid pressure operating beneath conventional military thresholds through simultaneous campaigns across informational, cyber, economic, and kinetic domains, including sabotage, infrastructure disruption, GPS interference, drone incursions, and coercive migration, with the fundamental vulnerability being institutional misalignment: Lithuanian decision-making continues to follow peacetime logic while Russia operates under wartime assumptions, probing vulnerabilities and measuring response capacity in ways that conventional defense frameworks are not structured to detect or counter.


Bajarunas identifies institutional adaptation velocity as Lithuania's primary vulnerability, the lag between threat evolution and institutional response cycles risks negating superior military capabilities, since decision-making speed and system-wide integration across military, infrastructure, civil society, and industrial domains matter more than platform-level capabilities, with Lithuania's 5.38% GDP defense allocation and German brigade deployment providing a capability baseline that structural gaps in cyber resilience, counter-drone coverage, and total societal mobilization currently undermine.


Source: Jamestown Foundation. Lessons Learned From Russia’s War Against Ukraine: The Case of Lithuania. [online] Published 18 August 2026. Available at: https://jamestown.org/lessons-learned-from-russias-war-against-ukraine-the-case-of-lithuania/


Kremlin Pivots from Climate Denial to Exploiting EU Policy Narratives

An analysis published by EUvsDisinfo states that Kremlin-aligned information operations have shifted from denying climate change to weaponising it as a vehicle for established anti-EU disinformation tropes, recasting EU climate and energy efficiency regulations as surveillance mechanisms and attacks on private property, amplifying apocalyptic predictions of European water wars and food price collapse to position Russia as a future climate haven, and repeatedly claiming that Europe faces winter without heating due to its rejection of Russian gas, while Russia's own fuel crisis driven by Ukrainian strikes on oil refineries and a diesel export ban imposed to stabilise domestic supply contradicts its narratives about European energy dependency.


The Kremlin's climate narrative pivot reflects a structural adaptation: as the observable effects of climate change became impossible to suppress domestically, with permafrost thaw threatening Russian oil and gas pipeline infrastructure, damaging buildings and railways, and the 2020 Norilsk fuel tank spill serving as an early indicator of accelerating technological risk, outright denial lost credibility with domestic audiences, so the adaptation shifted to attacking proposed solutions rather than denying the problem, a more durable posture that simultaneously exploits legitimate climate anxieties in target countries, reinforces anti-EU narratives without requiring false empirical claims, and allows Russia to reframe its fossil fuel dependency as a strategic advantage rather than an environmental liability.


Source: EUvsDisinfo. From Denial to Exploitation: How the Kremlin Has Adapted Its Climate Disinformation. [online] Available at: https://euvsdisinfo.eu/from-denial-to-exploitation-how-the-kremlin-has-adapted-its-climate-disinformation/


China

PRC Targets Kaohsiung Election with Fake Polling While AI Cyberattack Hits Taiwan Systems

An analysis published by AEI/ISW states that Taiwan's President Lai Ching-te publicly accused the PRC of targeting the Kaohsiung 2026 mayoral election with fabricated polling data and fake news, while a separate AI-enabled autonomous cyberattack in July 2026 compromised 21 government systems and exfiltrated approximately 2,500 records across 85 accounts, and Philippine President Ferdinand Marcos Jr. stated publicly that the Philippines would not allow its territory to be used as a US launch platform in a Taiwan contingency, a position the analysis links to ongoing PRC political influence on Manila.


The PRC's parallel deployment of disinformation and cyberattack operations against Taiwan reflects an integrated cognitive warfare doctrine in which election-targeted influence operations and infrastructure penetration serve complementary strategic functions; the fake polling and disinformation targeting Kaohsiung's 2026 vote serves as a rehearsal for 2028 presidential interference, while the AI-enabled cyberattack against government systems generates intelligence on official communications that can be used to calibrate or authenticate future disinformation content.


Source: American Enterprise Institute. China & Taiwan Update, August 18, 2026. [online] Published 18 August 2026. Available at: https://www.aei.org/commentary/china-taiwan-update-august-18-2026/


AI Enables PRC Comprehensive Election Interference Supply Chain Against Taiwan

An article published by Taipei Times states that Taiwan's Democratic Progressive Party China Affairs Department warned that artificial intelligence has significantly lowered the cost of producing disinformation, enabling the PRC to field a comprehensive election interference industry supply chain that collects voter data and online behavioral profiles, generates false audio recordings, videos, endorsements and scandals within hours before elections, distributes content through local collaborators and inauthentic accounts, and refines strategies based on real-time engagement metrics, operating at a speed calibrated to outpace Taiwan's fact-checking capacity.


An article published by Taipei Times states that the DPP's characterization of PRC election interference as an industry supply chain reflects a structural assessment that Beijing's interference is not episodic campaign activity but a continuous operational infrastructure evolved through documented iterations, the 2016 Chou Tzu-yu controversy, 2018 local election disinformation, 2020 presidential information warfare, and post-2022 influence operations, with each cycle providing feedback data that refines targeting precision, making the cumulative capability qualitatively different from a series of isolated incidents.


Source: Taipei Times. AI Could Intensify Chinese Election Interference: DPP. [online] Published 17 August 2026. Available at: https://www.taipeitimes.com/News/taiwan/archives/2026/08/17/2003862627


[AI Related Articles]


AI-Generated News Sites Now Indistinguishable from Authentic Outlets, NewsGuard Finds

A report published by NewsGuard Reality Check states that a growing infrastructure of AI-generated news websites, designed to pass as authentic local and national news outlets, publishing under credible-sounding names with content indistinguishable from genuine reporting to most readers, has proliferated to the point where professional media monitors cannot rapidly identify them without forensic analysis, illustrating that the barrier to creating authentic-appearing synthetic news infrastructure has fallen below the threshold that current detection systems and public media literacy effectively address.


The proliferation of AI-generated fake news sites represents a structural shift in the disinformation threat landscape; unlike fabricated stories distributed through identifiable partisan channels, AI-generated outlet networks establish apparent institutional credibility that makes individual false claims harder to debunk, since audiences attribute the authority of a news organization to content appearing under its masthead, and since AI generation enables the simultaneous operation of hundreds of such sites at negligible cost, the scale at which synthetic local news infrastructure can operate exceeds the monitoring capacity of national fact-checking organizations.


Source: NewsGuard's Reality Check. Is This Website Real or AI Slop? [online] Published 17 August 2026. Available at: https://www.newsguardrealitycheck.com/p/is-this-website-real-or-ai-slop


SOCOM Commander Warns AI Deception Has Made Officials Unable to Distinguish Real from Fabricated Content

An article published by Defense One states that Admiral Frank Bradley, Commander of US Special Operations Command, warned publicly that AI-enabled deception has made the information environment so opaque that even senior military and intelligence officials cannot reliably distinguish authentic from fabricated content, that counter-influence structures in the United States have been systematically dismantled, and that hostile active measures have in his words 'eaten the world', directly raising the probability of large-scale military conflict by increasing the likelihood that decision-makers will act on fabricated intelligence.


Admiral Bradley's public warning from a serving SOCOM commander is analytically significant beyond its substantive claims; the willingness of a sitting four-star commander to characterize the information environment in terms that directly contradict the stated confidence of political leadership signals that the gap between classified assessment and public communication on cognitive warfare vulnerability has reached a point where uniformed commanders judge public disclosure serves national security more than maintaining the appearance of institutional capability.


Source: Defense One. AI-enabled Deception Threatens Future Operations, Raises Chances of Large Conflict, Says Special Operations Leader. [online] Published 21 August 2026. Available at: https://www.defenseone.com/threats/2026/08/ai-enabled-deception-threatens-future-operations-raises-chances-large-conflict-says-special-operations-leader/415582/


LLM Political Persuasion Matches Human Content as Monitoring Frameworks Lag

A study published by Carnegie Endowment states that large language models have achieved political persuasiveness comparable to human-written content, with research across 19 models and over 700 political issues finding that factual claims drove persuasiveness equally between AI and human outputs, while a 2025 NewsGuard analysis found LLMs asserted false claims as true 35% of the time on misinformation queries, and the authors document that content moderation of politically sensitive topics has varied during live events, with GPT-4.1 refusal rates on Israel-related content increasing substantially during the August 2025 Gaza conflict without public disclosure of the change.


The authors identify the DEEP properties of LLMs, Dynamic (changing constantly without disclosure), Ephemeral (outputs disappear after use), Embedded (operating within layered software guardrails that vary by deployment), and Personalized (responses calibrated to individual user history), as the structural reason one-off audits are insufficient: a point-in-time audit captures a version of the system that may be materially different from what any user encounters at a given political moment, meaning LLMs can shift their political outputs between elections, between news cycles, and between individual users without any of those changes being detectable through current research or regulatory frameworks.


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


[General Reports]


Forthcoming Book Maps Authoritarian Disinformation Architecture and Proposes Multi-Layer Counter-Framework

A preprint book published by PhilPapers states that authoritarian regimes systematically deploy disinformation through layered propaganda architectures that exploit cognitive biases, social media recommendation algorithms, and AI-generation tools to sustain political control and undermine democratic institutions, with a typology spanning authoritarianism, polarization, propaganda mechanics, cognitive vulnerability exploitation, disinformation operations, and social media amplification, arguing that AI-generated deepfakes and coordinated inauthentic behavior represent a new operational tier that lowers fabrication costs while outpacing existing detection and regulatory frameworks, and that effective counter-IO requires simultaneous supply-side platform regulation and demand-side cognitive inoculation rather than relying on post-exposure fact-checking alone.


Friedenberg's interdisciplinary framework, integrating social psychology research on belief persistence and cognitive bias with structural analysis of social media platform architecture and authoritarian regime behavior, provides a practitioner-relevant taxonomy for classifying and anticipating disinformation operations, with its core insight being that the effectiveness of disinformation does not depend on the audience being irrational but on the operation being architecturally calibrated to exploit predictable cognitive processing patterns, including confirmation bias, illusory truth effects from repetition, and motivated reasoning, that are universal human features rather than individual vulnerabilities, meaning that counter-IO frameworks relying on media literacy alone address a symptom rather than the structural mechanism through which authoritarian actors convert information environments into instruments of political control.


Source: Fricker, Miranda. Disagreement and the State of Knowledge. [online] Available at: https://philpapers.org/rec/FRIDAD-6


SE Asia Insiders Confirm Shift from Bots to Data-Driven Gray Campaigns

A study published by Information, Communication & Society states that a peer-reviewed study based on 78 in-depth interviews with influence operation insiders in Indonesia, the Philippines, and Thailand identifies five simultaneous transformations in contemporary IO, a shift from state-dominated commissioning to a broader range of political and business clients; increased outsourcing to private PR firms and freelancers; a move from high-volume bot-like accounts to fewer carefully crafted pseudonymous influencer accounts with large genuine followings; a content shift from repetitive state-mirroring messaging to data-analytics-driven 'gray' campaigns that reframe real news rather than fabricate disinformation; and dissemination strategies moving from mass flooding to responsive, real-time, data-informed communication.


The IO 2.0 framework carries two analytically significant implications for counter-IO efforts: first, the shift from fabrication to gray campaigns, where real information is selectively reframed and amplified rather than invented, directly limits the effectiveness of fact-checking and content-removal moderation, since gray content does not trigger false-claim detection; second, the shift from automated bots to carefully nurtured pseudonymous accounts with genuine followings means contemporary IO is designed to be indistinguishable from ordinary user behavior at the account level, complicating coordinated inauthentic behavior detection frameworks that social media platforms rely on, a structural gap the authors conclude requires financial transparency of political campaigns and independent investigative journalism rather than technical content moderation alone.


Source: Ruijgrok, Kris, Berenschot, Ward, Sastramidjaja, Yatun, Gaw, Fatima, Sombatpoonsiri, Janjira, Wiyayanto, and Agonos, Mariam Jayne. Influence Operations 2.0: The Evolution of Social Media Manipulation in Southeast Asia. [online] Published 2026. Available at: https://www.tandfonline.com/doi/full/10.1080/1369118X.2026.2713663


Major Ad-Tech DSPs Place Own Brand Ads Next to Misinformation

An investigation published by NewsGuard Reality Check states that the four largest demand-side advertising platforms, Google's Display & Video 360, Amazon DSP, Yahoo DSP, and Adobe Advertising DSP, placed programmatic advertisements for their own brands and client brands next to misinformation articles advancing provably false claims, with NewsGuard analysts in 2026 identifying advertisements for 45 major technology brands appearing next to 162 false-claim articles, including Adobe Acrobat Pro ads appearing on JoeHoft.com next to false 2020 election theft articles and Google product advertisements appearing on a health hoax site next to articles claiming hydrogen peroxide cures cancer, exposing a fundamental gap between the brand safety guarantees these platforms sell to advertisers and the content adjacency their automated systems actually produce.


The inability of major ad-tech demand-side platforms to prevent their own advertising spend from appearing next to misinformation reveals the structural mechanism by which the disinformation ecosystem is commercially sustained: programmatic advertising operates through automated buying systems with limited publisher-level content visibility, meaning that misinformation sites monetize through the same commercial revenue channels as credible publishers, while the brand safety tools sold by these DSPs operate as opt-in domain blocklists rather than real-time content assessment, a design structurally incapable of addressing the velocity at which new misinformation sites and AI-generated content farms emerge, which means brand safety revenue subsidizes the commercial viability of maintaining a large, continuously replenishing ecosystem of misinformation sites.


Source: NewsGuard's Reality Check. These Ad-Tech Giants Claim to Protect. [online] Available at: https://www.newsguardrealitycheck.com/p/these-ad-tech-giants-claim-to-protect


Fake Polling Company Admits Fabricated US State Surveys Were a Social Experiment

An article published by Washington Times states that a polling company named Median Strategies admitted that fabricated surveys it published across at least three US states were conducted as a social experiment, with the company having no verifiable address, staff, or methodology disclosures, and the fabricated data having circulated in political media without verification before the admission,  illustrating how synthetic polling data can enter the political information ecosystem through normal political reporting channels in the absence of industry-standard source verification.


The Median Strategies case reveals a documented vulnerability in how political polling data is consumed and reported: survey data published under a professional-sounding organization name routinely receives less source verification than individual factual claims, because polling is understood as systematic measurement rather than an assertion subject to standard journalistic source assessment, creating a structural entry point for fabricated quantitative data into the political information environment that does not require the audience to accept a specific false claim, only a numerical framing as reflecting public opinion.


Source: The Washington Times. Fake Polling From Mysterious Company Highlights Danger of Unvetted Election Surveys. [online] Published 18 August 2026. Available at: https://www.washingtontimes.com/news/2026/aug/18/fake-polling-mysterious-company-highlights-danger-unvetted-election/ 


US Lawmakers Demand Meta Detail AI Deepfake Election Safeguards Ahead of Midterms

An article published by India West states that 19 US lawmakers led by Representative Kevin Mullin sent a letter to Meta CEO Mark Zuckerberg demanding specific details of the company's election safeguards against AI-generated deepfakes ahead of the 2026 midterm elections, citing documented rollbacks of Meta's previous election integrity commitments and the absence of publicly disclosed standards for detecting or removing AI-generated political impersonation content from Facebook and Instagram.


The bipartisan letter to Zuckerberg reflects a shift in Congressional strategy from legislative proposals, which have failed to advance in multiple sessions, to oversight pressure using platform commitments already made as the accountability standard, rather than demanding new regulatory obligations, the letter specifically asks Meta to account for the safeguards it has already committed to and explain where those commitments have been reduced, placing the burden of justification on the platform rather than requiring legislation to establish new obligations.


Source: India-West. 19 US Lawmakers Slam Meta Over AI Deepfakes Ahead of Midterms. [online] Published 18 August 2026. Available at: https://indiawest.com/19-us-lawmakers-slam-meta-over-ai-deepfakes-ahead-of-midterms/


AI Chatbots Fill Voter Information Gap as CISA Guidance Links Break Before Midterms

An analysis published by TechPolicy.Press states that AI chatbots are becoming a primary source of electoral information for an increasing share of voters ahead of the 2026 midterm elections, that political campaigns are using synthetic voter focus groups, AI-simulated representations of voter segments, to craft and test messaging, and that official CISA guidance links to election information resources have broken following structural changes to the agency's public communications, leaving voters who seek authoritative federal information directed to dead links.


The convergence of AI-delivered voter information and broken CISA guidance links creates a structural vacuum: voters seeking authoritative election information from federal sources encounter broken links while AI chatbots, which the Brennan Center documented as making factual errors in 50% of election-related responses despite rebuffing conspiracy theories,  fill that informational space, meaning the practical effect of degraded government communications infrastructure is to transfer voter guidance from verified official sources to AI systems whose accuracy at the task has been documented as unreliable.


Source: Tech Policy Press. AI Meets the US Midterm Elections. [online] Published 16 August 2026. Available at: https://www.techpolicy.press/newsletter-august-16-2026/


[Appendix - Frameworks to Counter Disinformation]


FTC Expands Operation AI Comply to B2B with Personal Liability

An analysis published by Holland & Knight states that the Federal Trade Commission's Operation AI Comply enforcement campaign, launched in September 2024, has expanded from initial B2C consumer protection cases to target business-to-business AI capability misrepresentation, using the means and instrumentalities doctrine to hold corporate decision-makers personally liable alongside their companies, while the Securities and Exchange Commission has launched a parallel AI washing enforcement campaign targeting publicly traded companies that misrepresent their AI capabilities to investors.


The expansion represents a structural progression from consumer protection framing, where the primary harm is individual consumer deception, to systemic market integrity framing, in which false AI capability claims distort investment decisions, competitive positioning, and procurement outcomes across entire industry sectors, with personal liability exposure for corporate officers providing a deterrent that company-level fines alone may not achieve.


Source: Holland & Knight. “Operation AI Comply” 2 Years Later: Continued Enforcement Against Misleading Claims. [online] Published 18 August 2026. Available at: https://www.hklaw.com/en/insights/publications/2026/08/operation-ai-comply-2-years-later-continued-enforcement


[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.










 
 
bottom of page