_edited.png)
.png)
Cyber based influence campaigns 12th – 18th October 2026 Report
[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 12th to the 18th of October 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]
X
Tiktok
Russia
China
Iran
[ Report Highlights]
X said it suspended more than 13,000 accounts in suspected Chinese, Russian, and Iranian influence operations targeting US politics, without publishing supporting evidence.
Bloomberg, cited by Kyiv Post, found Russia's Matryoshka operation targeting Democrats in 20 of 35 Senate races, while researchers documented its AI fakes and bot-inflated views.
The head of MI5 confirmed that Russia tried to interfere in the 2016 Brexit referendum but said the attempts were neither particularly sophisticated nor successful.
DFRLab traced four coordinated networks of Facebook pages, websites and Telegram channels that amplified war and resettlement fears before Armenia's June 2026 election.
RSF found that pro-Kremlin narratives dominated the Ukraine-war videos it monitored on Bilibili, reaching Chinese Gen Z through bloggers and Russian state media.
NewsGuard showed how a deceptively edited phone call fuelled a false claim, viewed millions of times, that Senate candidate James Talarico was in a mental facility.
CRC found a Russia-linked network, UNK_SmoothOperator, combining propaganda with attempts to recruit Europeans as drone operators and calls for sabotage against NATO targets.
[ Report Summary]
X said it suspended more than 13,000 accounts in suspected influence operations linked to China, Russia and Iran that targeted US political debate before the midterms.
An AI-generated TikTok persona resembling actress Jenna Ortega gained about one million followers and reportedly stopped labeling its videos as AI-made, Cybernews reports.
The head of MI5 said Russia tried to interfere in the 2016 Brexit referendum via RT, Sputnik and trolls, but that the attempts were not particularly successful.
Kyiv Post, citing Bloomberg, reports that the Russian Matryoshka operation used AI fakes and bot-inflated view counts against Democratic candidates before the midterms.
A Carnegie analysis argues that Russia's fragmented AI propaganda has limited ability to rewrite chatbot answers but exploits data voids and mimics genuine debate.
DFRLab found four coordinated networks of Facebook pages, websites and Telegram channels amplifying pro-Russian and opposition fear narratives before Armenia's June 2026 election.
Pro-Kremlin accounts and Russian media spread AI-generated images falsely suggesting Ukraine is drafting or luring women into the army, the Kyiv Independent reports.
RSF found that pro-Kremlin narratives dominated the Ukraine-war videos it monitored on Bilibili, spread by popular Chinese bloggers and Russian state media.
Recorded Future assesses that the Kremlin has escalated VPN blocking, managed shutdowns and pressure on Telegram and WhatsApp while pushing the state-owned Max messenger.
An official familiar with China's united front tactics warned that Beijing would likely run a disinformation campaign in the Taipei mayoral race, possibly using deepfakes and bots.
Jamestown reports that China is trying to isolate Japanese Prime Minister Takaichi by courting her domestic rivals, invoking UN 'enemy state' clauses and probing US-Japan ties.
A Simon Wiesenthal Center report argues that Iran has used antisemitism for three decades as a tool to divide American society and erode trust in institutions.
Israeli election officials will run an influence-operations war room with major platforms on election day amid reported Iranian and Russian campaigns targeting voters.
BOOM found that social media users in India shared incorrect Grok and Gemini replies to falsely dismiss genuine photographs of police crackdowns on recent protests as old.
Malaysia's communications regulator is seeking more information from Anthropic after the company reported that Claude was used in a foreign-linked operation targeting Malaysian users.
CBS News Texas found that a Facebook page with apparently fake contact details posted AI-generated 'documentaries' of the 2009 Fort Hood shooting riddled with factual errors.
VisionHeight found a network of hyperlocal websites across at least 14 US states apparently built to influence how AI chatbots answer questions about political candidates.
NewsGuard found that conservative accounts used a deceptively edited phone call to falsely claim Texas Senate candidate James Talarico was treated at a mental health facility.
NewsGuard's False Claim of the Week debunks viral posts claiming that ballots marked by poll workers are disqualified, a claim with 2.4 million views across 614 posts.
NewsGuard found that a Republican PAC launched MichiganderNews.com, a site posing as independent local news while attacking Democratic Senate candidate Abdul El-Sayed.
The Kyiv Independent found no data showing that Ukrainian strikes on Yandex data centers disrupted Russian influence operations, despite viral claims and official statements.
Officials told Nextgov/FCW that federal cooperation against foreign influence has weakened before the midterms, even as AI makes influence operations cheaper to run.
CRC found that a Russia-linked cross-platform network it calls UNK_SmoothOperator combined propaganda with attempts to recruit Europeans as drone operators and calls for sabotage.
The UK Electoral Commission is developing an AI system that aims to alert candidates to significant deepfakes within three hours of posting, its chief executive said.
Austria's foreign minister announced a Center of Excellence Against Foreign Information Manipulation, modelled on Helsinki's Hybrid CoE and due to launch in spring 2027.
A WashU and Caltech study of more than 4,200 US voters found that AI-generated pre-bunking articles reduced belief in false election rumours, with a smaller lasting effect.
[State Actors]
X
X Suspends More Than 13,000 Accounts Over Foreign Influence
An article published by Voice of America (VOA) states that X said it has suspended more than 13,000 accounts involved in suspected foreign influence operations linked to China, Russia, and Iran that targeted U.S. political discourse ahead of the November 3 midterm elections. In a statement posted on October 14, the platform said the networks used inauthentic accounts to post and amplify political content targeting candidates and public figures across the political spectrum.
VOA notes that the company did not publicly provide evidence supporting its findings or specify how many of the suspended accounts were linked to each country. X also pointed to its civic integrity policy, which covers content that could confuse voters about how, when, or where to participate or discourage participation, and said users can report such posts under a 'Civic Integrity' category, with enforcement continuing through Election Day and beyond.
Source: Voice of America (VOA). X Suspends 13,000 Accounts Linked to China, Russia and Iran. [online] Published 16 October 2026. Available at: https://www.voanews.com/a/x-suspends-13-000-accounts-linked-to-china-russia-and-iran-10162026/8212051.html
Tiktok
AI-Generated Jenna Ortega Lookalike Draws a Million TikTok Followers
An article published by Cybernews states that a TikTok account called Friday Adams, which uses an AI-generated lookalike of actress Jenna Ortega, has roughly 1 million followers, and that one 15-second video reportedly drew 67.6 million views and may be worth about $27,000. Despite telltale signs, such as changes to the persona's appearance depending on the angle, users still believe the account is the real Ortega, and its bio urges followers to 'subscribe for more talk with me'.
For a time, the account labeled its content as AI-generated. Still, its newer videos reportedly no longer carry the label, which Cybernews says conflicts with TikTok's rules requiring creators to label realistic synthetic images, audio and video and prohibiting account behaviour that may 'spam or mislead our community'. The account was still active at the time of writing; the X account Kelmruns, which Cybernews identifies as the original source, claims the persona was generated with an OpenAI model, and Cybernews said it had contacted TikTok and Ortega's management for comment.
Source: Cybernews. Jenna Ortega Deepfake Videos Spread on TikTok. [online] Available at: https://cybernews.com/ai-news/jenna-ortega-deepfake-tiktok/
Russia
MI5 Chief Confirms Russian Attempt to Interfere in Brexit Vote
An article published by The Guardian states that Ken McCallum, head of MI5, said in his annual threat update that Russia did try to interfere in the 2016 Brexit referendum through social media and television reports, endorsing comments made by Prime Minister Andy Burnham. McCallum said Russian-owned broadcasters RT and Sputnik 'came in pretty strongly around the time of the referendum' with biased anti-migration and anti-EU reports, and that there had been 'a second line of interference attempted through trolls', an unspecified number of fake accounts operating largely on X.
McCallum said neither tactic was particularly sophisticated or successful and warned that prolonging the debate risked handing the Kremlin a needless propaganda victory, since continued attention gives the impression that its disinformation is more effective than it is. Downing Street said it had been confirmed that Russia interfered in the 2019 general election and that it was 'inconceivable that they didn't in 2016', while adding that it did not believe the interference was successful in the referendum.
Source: The Guardian. Russia Interfered in Brexit Referendum, MI5 Warned. [online] Published 13 October 2026. Available at: https://www.theguardian.com/politics/2026/oct/13/russia-brexit-eu-referendum-interference-mi5
Russian Matryoshka Bots Inflate Fake Videos Targeting US Midterms
An article published by Kyiv Post states that, according to Bloomberg News, Russian state-backed operators are using automated bot networks to inflate view counts of fabricated videos targeting Democratic candidates and Hollywood celebrities ahead of the US midterm elections, in a campaign researchers call 'Matryoshka' or 'Operation Overload'. A Bloomberg review of 130 fabricated videos and doctored newspaper front pages circulated between September 10th and October 9th, 2026, found a campaign focused almost exclusively on Democratic contenders, including a video with a counterfeit ABC News banner claiming Texas Senate nominee James Talarico paid OnlyFans models, which his spokesperson denied.
The findings were documented by the anonymous research group Antibot4Navalny and corroborated by Graphika and dTeam: researchers observed that Matryoshka posts can jump to as many as 200,000 views within the first hour before quickly plateauing, while the amplifying accounts had been dormant for years, had few followers and generated almost no genuine reposts or likes. According to Bloomberg's analysis, the network has moved beyond presidential contests to down-ballot races, targeting Democratic candidates in 20 of the 35 Senate races and 29 of the 36 gubernatorial contests on the November 3 ballot.
Source: Kyiv Post. Russian ‘Matryoshka’ Bot Farm Targets US Midterms With AI. [online] Published 18 October 2026. Available at: https://www.kyivpost.com/post/87423
Kremlin's Chaotic AI Propaganda Mimics Genuine Public Debate
An analysis published by the Carnegie Endowment for International Peace states that Russia is trying to influence the answers of global AI chatbots, primarily through the material they find on the internet, but that the effectiveness of this 'LLM grooming' remains unclear. Author Ilya Klishin notes that the American Sunlight Project counted more than 20,000 posts across 97 websites of the Pravda network in 48 hours, and that while NewsGuard stress tests found chatbots repeating Russian false claims in about a third of responses to test queries, many of them leading, a peer-reviewed HKS Misinformation Review study of 416 responses found that 5 percent supported a false claim and that links to propaganda sites appeared mainly in answers to niche questions with few reliable sources, which its authors attributed to 'data voids'.
Klishin argues that the fragmented, often contradictory output of RT, Sputnik, intelligence agencies, private contractors and Prigozhin-era projects is not necessarily a problem for the Kremlin, because the resulting cacophony resembles genuine public debate and can lead a chatbot that fails to recognise a common source to present facts and fabrications as equally valid versions, part of what he calls 'effective inefficiency'. He also cautions that the system has limits that have likely already been reached or nearly reached, since each generation of models will be more resilient, and that the threat should not be blindly extrapolated to 2030, let alone 2035.
Source: Carnegie Endowment for International Peace. In the Kremlin’s AI Propaganda Blitz, the Noise Is the Signal. [online] Published 6 October 2026. Available at: https://carnegieendowment.org/russia-eurasia/politika/2026/10/russia-AI-political-influence
Coordinated Networks Amplified War Fears Before Armenia's Election
An investigation published by the Digital Forensic Research Lab (DFRLab) states that four networks of coordinated, often inauthentic Facebook pages and websites, together with Telegram channels posting in synchronization, amplified claims ahead of Armenia's June 7th, 2026, parliamentary election that hundreds of thousands of Azerbaijanis would come to Armenia and that the vote would bring war. The more than one hundred pages hid behind entertainment-style names such as 'International Beauties' or impersonated news outlets, posted identical content within minutes or seconds, leaned toward opposition and pro-Russian positions, and in several cases stopped posting on June 9, two days after the vote.
The investigators linked assets through technical signals: four fringe websites had near-consecutive Yandex Metrica IDs suggesting they were set up in one batch, several Russia-hosted sites shared a server IP address, and Russian-language admin traces, and a former page developer whose email appears on some first-network pages told FIP that in mid-2025 he sold about 100 pages with three to four million followers to buyers he said were based in Russia. On Telegram, two channel pairs each posted the same object within a minute more than 500 times between May 8th and June 14th, often in the same second, a pattern the researchers say suggests one operator may run both channels in each pair.
Source: Digital Forensic Research Lab (DFRLab). How Coordinated Inauthentic Networks Amplified Fears of War and National Loss Before Armenian Elections. [online] Published 12 October 2026. Available at: https://dfrlab.org/2026/10/12/inauthentic-networks-amplify-fears-of-war-before-armenian-elections/
Russia Uses AI Images to Fake a Women's Draft in Ukraine
A fact-check published by The Kyiv Independent states that pro-Kremlin social media accounts and Russian media have spread AI-generated images alleging that Ukrainian women are being lured into the Armed Forces by recruitment posters or outright called up for service, although no mandatory draft of women is in place or being prepared. A September 15th post claimed such posters had appeared around Odesa while other channels placed the same photographs in Lutsk; LetsData's Roman Osadchuk said 'there were no such banners' and that an AI verification tool indicated they were AI-generated, and a similar image posted by Russian state TV presenter Olesya Loseva was rated 99.9 percent likely AI-generated when the Kyiv Independent ran it through Hive Moderation.
The campaign emerged as rumours of women's conscription circulated in Ukraine, after former Defense Minister Oleksii Reznikov argued on August 20th that requiring women to serve would 'immediately double the mobilization pool' and former U.S. envoy Keith Kellogg said on August 24th that women should be mobilized. About 76,000 women serve in Ukraine's Armed Forces as volunteers, and women with medical or pharmaceutical education must be on the military register, which does not mean they can be mobilized against their will; Osadchuk describes the narrative as recurring and periodically amplified with 'new tools' such as AI imagery.
Source: The Kyiv Independent. Fact Check: Russia Turns to AI to Push Fakes About Women’s Draft in Ukraine. [online] Available at: https://kyivindependent.com/fact-check-russia-turns-to-ai-to-push-fakes-about-womens-draft-in-ukraine-2/
Kremlin War Narratives Reach China's Gen Z Through Bilibili
An investigation published by Reporters Without Borders (RSF) states that, in monitoring more than a hundred videos about the war in Ukraine on Bilibili, the Chinese video platform popular with Gen Z, it found that the large majority reflected pro-Kremlin narratives, including mockery of Zelensky and Ukrainian mobilisation and justifications for the invasion. The content is driven by popular Chinese bloggers such as Lu Kewen, whose most-viewed video on the invasion had more than 11 million views by summer 2026, and by RT and Sputnik, which are almost the only foreign state media officially present on the platform, with 1.3 million and 628,000 followers respectively.
RSF says graphic violence downloaded from Russian pro-war Telegram channels is popular despite Bilibili's ban on gore, with videos showing the bodies of Ukrainian soldiers drawing up to two million views, and that Chinese bloggers and Russian state media continue to profit from the war on Bilibili, where a creator incentive programme pays for engagement. The organisation concludes that Bilibili has become an increasingly significant channel for Russia and that 'once again, Beijing and Moscow work together on information warfare'; Bilibili did not respond to repeated requests for comment, and RT declined to answer and called RSF 'sinister'.
Source: Reporters Without Borders (RSF). Selling the Kremlin’s War Narrative to China’s Gen Z: Inside Platform Bilibili’s Coverage of Ukraine. [online] Published 15 October 2026. Available at: https://rsf.org/en/selling-kremlin-s-war-narrative-china-s-gen-z-inside-platform-bilibili-s-coverage-ukraine
Kremlin Escalates VPN Blocks and Pushes State Messenger Max
A report published by Recorded Future states that, since mid-2025, the Kremlin has intensified restrictions on digital communications and infrastructure, which its Insikt Group assesses are very likely aimed at suppressing the organisation of domestic political dissent and conducting digital surveillance. Roskomnadzor had blocked 258 VPN services by October 2025 and 460 by early 2026, and in 2025 independent trackers recorded more than 37,000 cumulative hours of blackouts affecting nearly 80 percent of Russia's territory, with 74 of 83 regions introducing allowlists of government-approved sites.
The report says the Kremlin moved to fully block WhatsApp by February 2026, heavily throttled Telegram, and mandated from September 2025 that the state-owned Max messenger be pre-installed on all devices sold in Russia and Belarus, while independent security analysts assess that Max lacks end-to-end encryption for chats and shares user metadata with intelligence and law enforcement agencies. Insikt Group nonetheless judges a fully 'sovereign RuNet' unlikely in the near term, noting that VPN use has grown to 39 percent of Russians according to Kommersant and that Max remains unpopular.
Source: Recorded Future. Russia’s Restrictive Information Environment. [online] Available at: https://www.recordedfuture.com/research/russia-restrictive-info-environment
China
Taiwan Official Warns of Chinese Disinformation in Taipei Mayoral Race
An article published by the Taipei Times states that an official familiar with China's 'united front' tactics said Beijing would likely try to interfere in the Taipei mayoral election through a disinformation campaign aimed at shaping public perception, because it has become more difficult to work through Chinese immigrants. The official said China could use deepfake videos to mislead people and target specific candidates, that Beijing is to devote substantial resources to smearing candidates it does not like as the election approaches, and that large volumes of disinformation could spread through deepfakes and bot accounts.
In August, Chinese immigrant Xu Chunying was sentenced to seven years in prison for fraud, helping Chinese nationals enter Taiwan illegally and acting at Beijing's direction to influence Taiwanese politics, and the Cross-Strait Marriage and Family Service Alliance, where she was secretary-general, was dissolved in May, steps the official said would make it harder for Beijing to mobilise Chinese spouses. With few Chinese immigrants running this year, the official argued that the main threat authoritarian states pose to democratic elections is cognitive warfare over a short period rather than organised voting by local groups, adding that anyone mobilised to interfere would be investigated under the law.
Source: Taipei Times. Official warns of Chinese disinformation campaign to influence Taipei election. [online] Published 12 October 2026. Available at: https://www.taipeitimes.com/News/taiwan/archives/2026/10/12/2003865838
Beijing Campaigns to Isolate Japan's Prime Minister Takaichi
An analysis published by the Jamestown Foundation states that, nearly a year into its diplomatic dispute with Japan, the People's Republic of China has shifted from economic pressure to a multi-faceted campaign to isolate Prime Minister Sanae Takaichi over her pro-Taiwan stance at home, in the United Nations and within the U.S.-Japan relationship. Beijing has hosted delegations of Japanese politicians it views as friendly, including former Foreign Minister Takeshi Iwaya, while its foreign ministry accuses Japan of 'new-type militarism' and the existence of the UN Charter's obsolete 'enemy state' clauses serves Beijing's anti-Japan propaganda.
The analysis also describes Beijing seizing on a White House fact sheet recalling that the United States and China were World War II allies, which Xinhua's English translation rendered as a 'China-U.S. joint fight against Japanese aggression'. Jamestown judges the effort unlikely to succeed because Beijing's preferred partners in Tokyo lack political clout and the U.S.-Japan alliance remains strong, and says the campaign looks like an attempt to punish Takaichi for challenging a PRC 'red line' and to make an example of her to other countries.
Source: Jamestown Foundation. PRC Launches Campaign to Isolate Japan. [online] Published 14 October 2026. Available at: https://jamestown.org/prc-launches-campaign-to-isolate-japan/
Iran
Wiesenthal Center Report Maps Iran's 30-Year Influence Campaign
An article published by The Jerusalem Post states that a report released by the Simon Wiesenthal Center (SWC) argues that a decades-long Iranian campaign has amplified antisemitism and anti-Israel rhetoric across the Western political spectrum to erode trust in American institutions and forge alliances across ideological divides. The report says that by 2026 Iran's influence operations had expanded to fake news websites mimicking American media, fabricated personas posing as American activists, campaigns simulating grassroots support, and manipulation of Wikipedia, and it identifies more than 340 influencers, outlets, organisations and affiliated entities that it says form an interconnected influence network.
The SWC describes a six-stage cycle that runs from caveated reports by humanitarian organisations and academic framing, through alternative outlets that strip the caveats and add inflammatory framing and commentators who repackage the content for left- and right-wing audiences, to activist organising and conspiracy theorists who cite those outlets as evidence. Its recommendations include mandatory coordinated-threat disclosure by platforms, mandatory labelling of AI-generated content and a congressional review of Section 230 to remove immunity for content attributable to designated foreign adversaries or their proxies.
Source: The Jerusalem Post. Article title unavailable. [online] Available at: https://www.jpost.com/diaspora/antisemitism/article-911698
Israeli Officials Set Up War Room Against Foreign Election Influence
An article published by World Israel News states that Din Livne, acting director-general of Israel's Central Elections Committee, said an 'influence operations war room' would operate on the October 27th election day in coordination with Meta, Microsoft, Google and other major platforms, and that campaigns intended to undermine the proper conduct of the vote 'will be removed immediately'. Shin Bet chief David Zini told lawmakers in a closed Knesset briefing in July that Iran was the main foreign actor seeking to interfere in the election, with Russian and Chinese activities also reportedly discussed.
Officials and researchers cited in the report say the objective is not necessarily to change votes but to weaken public trust through disinformation, online manipulation and cyberattacks. The open-source research group Golden Owl reported that fake Israeli personas on X linked to Iran generated about 950,000 views in 30 days and identified 237 suspicious accounts, while the watchdog FakeReporter said most of the operations it identified were linked to Iran, with others attributed to Russia, and that one image circulated by Netanyahu supporters as evidence of incitement was traced to an account apparently connected to an Iranian operation.
Source: World Israel News. Israel Braces for Foreign Influence Operations Ahead of Oct. 27 Election. [online] Published 17 October 2026. Available at: https://worldisraelnews.com/israel-braces-for-foreign-influence-operations-ahead-of-oct-27-election/
[AI Related Articles]
Chatbot Replies Wrongly Dismiss Genuine Indian Protest Photos as Old
A fact-check published by BOOM states that social media users are sharing screenshots of Grok and Gemini replies that wrongly describe real photographs of police crackdowns during recent protests in Delhi and other parts of India as old and unrelated. A photo of a woman detained with a child, taken by Hindustan Times photojournalist Raju Shinde on October 7th, 2026, during a tribal student protest in Mumbai, was 'fact-checked' with a Gemini reply dating it to a 2024 protest, while Grok wrongly placed an October 11th, 2026 New York Times image of a police officer appearing to choke a Sikh man in New Delhi at a 2020 protest in Kolkata.
When BOOM tested the chatbots itself, Gemini gave yet another wrong origin, placing the Mumbai image in a 2023 incident in Kerala on the basis of an X post with no evidence, and Grok repeated its incorrect answer while citing irrelevant sources. The error spread beyond social media when NDTV used the image and Gemini's response in a now-deleted fact-check video, and BOOM warns that because Grok's responses often reflect the narratives circulating on X, using chatbots for fact-checking during breaking news risks amplifying false narratives.
Source: BOOM. Incorrect Grok and Gemini Answers Used to Discredit CEC Protest Images. [online] Available at: https://www.boomlive.in/fact-check/incorrect-grok-and-gemini-answers-used-to-discredit-cec-protest-images-32367
Malaysia Questions Anthropic Over Alleged Claude-Linked Influence Operation
An article published by Scoop states that the Malaysian Communications and Multimedia Commission (MCMC) is seeking further information from Anthropic over allegations that its AI technology was used in a foreign-linked operation to manipulate political influence targeting Malaysian users. According to the Communications Ministry's parliamentary reply, Anthropic alleged that a foreign entity used Claude to develop several components of the operation, including targeting systems, networks of fake social media accounts and fake news websites, and linked the operation to BBS Bilisim Teknolojileri, an Istanbul-based technology company.
The ministry said Anthropic reported terminating the identified Claude accounts and implementing mitigation measures, and that any regulatory or enforcement action would follow under the Communications and Multimedia Act 1998 and the Online Safety Act 2025 once sufficient information and evidence had been obtained. It added that the Digital Ministry is drafting an AI Governance Bill covering accountability, risk management and the responsibilities of those who develop and implement AI systems.
Source: Scoop. MCMC Seeks Answers from Anthropic over Alleged AI-Driven Political Manipulation in Malaysia. [online] Available at: https://www.scoop.my/news/300746/mcmc-seeks-answers-from-anthropic-over-alleged-ai-driven-political-manipulation-in-malaysia/
Unverified Facebook Page Spreads Error-Filled AI Videos of Fort Hood Shooting
An article published by CBS News states that a Facebook page called 'Why Is It Like That?' posted a two-part series of AI-generated videos purporting to depict the 2009 Fort Hood shooting, which left 13 people dead, and that both are riddled with mistakes and falsehoods. The second video shows firefighters entering the building while gunfire was underway, although the first responders were police officers who encountered the shooter outside, and omits Sgt. Mark Todd, the civilian police officer who ultimately ended the shooting.
CBS News Texas could not verify who runs the page: its listed Austin address belongs to a tattoo and piercing shop whose employee had never heard of it, its phone number belongs to a home remodelling company, and an email to its address was returned as undeliverable. Both videos are labelled as 'true crime documentaries', one is captioned as intended for 'education, documentary and historical analysis purposes', and they have been viewed and shared numerous times, and CBS describes them as a snapshot of the many AI-generated depictions now circulating online.
Source: CBS News. AI-Generated Videos Spread Misinformation About Fort Hood Shooting. [online] Available at: https://www.cbsnews.com/texas/news/fort-hood-shooting-ai-generated-videos-misinformation/
Hyperlocal Website Network Seeks to Steer AI Chatbot Election Answers
An article published by CyberScoop states that research by Israeli threat intelligence firm VisionHeight found networks of hyperlocal websites, with names such as 'Wolverine State Report' and 'Peach State Bulletin', across at least 14 U.S. states, filled with statements such as 'Candidate X is the stronger choice' repeated hundreds of times with slight variations for different political issues. The content appears to be written for AI crawlers rather than humans; the sites were configured to allow major U.S. AI companies to scrape them, and VisionHeight confirmed that Perplexity cited one page, Tarheel State Report, as its first source for an answer.
VisionHeight's initial blog counted six states and 37 websites, all registered in a single evening, and 36 of them within an eight-minute window, and on September 30 the network published roughly 100,000 words of candidate content using the same template; CyberScoop notes it appears AI was used to generate much of the infrastructure and content. The firm does not attribute the activity to a particular actor, and co-founder Nitsan Daniel argues that such operations are easily reproducible by foreign and domestic actors and that AI makes what were once small targets more attainable.
Source: CyberScoop. The Next Election Influence Campaign Will Target Your AI Chatbot. [online] Published 15 October 2026. Available at: https://cyberscoop.com/ai-chatbots-election-influence-hyperlocal-networks/
[General Reports]
Deceptively Edited Call Fuels False Talarico Mental Hospital Claim
A report published by NewsGuard states that conservative social media accounts claimed Texas Senate candidate James Talarico checked into the Lindner Center of Hope, a mental health facility in Ohio, citing a recorded phone call with a facility employee that had been deceptively edited to remove the employee's repeated denials that Talarico was a patient. Influencer Rhein Amacher's post of the recording received 1.3 million views, and influencer Nick Sortor's post of it received 7.3 million views, after Talarico took a nine-day break from campaigning that he later attributed to the flu.
The Lindner Center said Talarico 'has never received treatment' there, and in the full recording, published by The Bulwark, the employee says she found no one by that name discharged in the previous four days; NewsGuard adds that Talarico appeared at the Austin City Limits festival on October 9th, the day before the call was supposedly made. Amacher told a right-leaning publisher that he initially gave the audio to The Blaze and the Daily Mail, which raised legal concerns, before giving it to influencers who are 'not held to the same journalistic standards'.
Source: NewsGuard Reality Check. Talarico’s Phantom Mental Breakdown. [online] Available at: https://www.newsguardrealitycheck.com/p/talaricos-phantom-mental-breakdown
False Claim Says Poll-Worker Markings Disqualify Ballots
A report published by NewsGuard states that social media users are falsely claiming that ballots with written markings from poll workers, such as a checkmark, letter or star, are disqualified in U.S. elections, a claim that garnered 2.4 million views across 614 posts and has re-emerged ahead of the November midterms, mostly on left-leaning accounts. A NewsGuard review identified no counties or states that automatically disqualify ballots because of such markings, and several states, including North Carolina, actually require poll workers to write on ballots.
NewsGuard notes that the claim has circulated for six years and was previously spread by both left- and right-leaning users, suggesting it may stem from a misunderstanding of voting procedures rather than partisan intent, yet it could still be used to raise doubts about election results. One widely shared post, screenshotted from a liberal Facebook group, urged voters to request a new ballot if anything was written on theirs, and a Threads repost on October 11th received 248,000 views.
Source: NewsGuard Reality Check. Marked-Up Ballots Are Okay. [online] Available at: https://www.newsguardrealitycheck.com/p/marked-up-ballots-are-okay
Republican PAC Launches 'Pink Slime' News Site in Michigan
A report published by NewsGuard states that the Senate Leadership Fund, a Republican political action committee, has launched MichiganderNews.com, a website presenting itself as an independent news source under the motto 'Trusted reporting for the state of Michigan' while consistently criticising Democratic Senate candidate Abdul El-Sayed and boosting Republican Mike Rogers. The only disclosure is a small 'Paid for by SLF PAC' note at the bottom of its pages, with no 'About' page explaining ownership.
NewsGuard counts the site as the 22nd new 'pink slime' outlet it has identified this year, bringing the national total to 1,181, and notes that the Courier network, backed by liberal groups, has run GanderNewsroom.com in Michigan since 2020. Meta advertising records show the PAC spent an estimated $59,400 to $76,815 on Facebook and Instagram ads promoting the site's content, on top of $71,400 to $101,346 for a similar site targeting Roy Cooper in North Carolina; neither the PAC nor the site responded to requests for comment.
Source: NewsGuard Reality Check. GOP-Funded Pink Slime Targets Michigan. [online] Available at: https://www.newsguardrealitycheck.com/p/gop-funded-pink-slime-targets-michigan
No Evidence Yandex Strikes Crippled Russian Influence Operations
A fact-check published by The Kyiv Independent states that, after Ukrainian drones struck Yandex data centers in three regions between October 9th and 11th, viral claims of a sudden drop in Russian bot activity on social media were not supported by data. The confusion stems largely from a Cloudflare chart showing bot activity from Yandex's network falling to 0.3 percent, which measured automated web requests mainly from Yandex's own search crawlers rather than social media bots or trolls, and Poland's NASK Disinformation Analysis Center recorded no significant decrease in activity by disinformation accounts between September 30th and October 11th.
The claims were nonetheless echoed by officials, including Polish minister Tomasz Siemoniak, who claimed to have received security-service information about a 'radical decline' that has not been published, and Czech MEP Tomas Zdechovsky, who shared a graph purportedly linking the strikes to a drop in AfD-related activity that, according to Rohtext, lists no source. Russian disinformation is believed to rely heavily on decentralised proxy networks abroad, and researcher Dave Troy warns that overstating the strikes' effect is itself a strategic risk because 'the threat lies elsewhere and must still be assessed and fought'.
Source: The Kyiv Independent. No, Ukraine’s Strikes on Yandex Data Centers Didn’t Cripple Russian Influence Operations. [online] Available at: https://kyivindependent.com/no-ukraines-strikes-on-yandex-data-centers-didnt-cripple-russian-influence-ops/
Officials Report Weaker US Federal Cooperation Against Foreign Influence
An article published by Nextgov/FCW states that the Trump administration's retreat from programmes that track foreign influence operations has weakened cross-sector relationships used to identify such campaigns ahead of the midterm elections, according to current and former federal officials, state election officials and specialists, several of whom spoke anonymously. State officials say they receive fewer election-specific federal briefings than in previous cycles, a cooling that sources say began under the Biden administration; the Trump administration has since shuttered CISA's election disinformation work, disbanded the FBI's Foreign Influence Task Force and closed the State Department's office for countering foreign information threats.
Specialists say AI now helps operators write and distribute posts, study audience responses and adjust campaigns at scale, and one current official says Russia is more likely to pursue obvious inflammatory messaging this cycle while China's efforts may be more covert, while pro-Iran outfits have used AI more overtly in recent months. The same official cautions that more convincing content does not necessarily change how people vote, while CISA says it will activate an Incident Coordination Group around Election Day and the administration says ODNI has assigned officials to coordinate election threat intelligence.
Source: Nextgov. Officials See Gaps in Election Communications as AI Makes Foreign Influence Easier. [online] Published October 2026. Available at: https://www.nextgov.com/cybersecurity/2026/10/officials-see-gaps-election-communications-ai-makes-foreign-influence-easier/416583/
Russia-Linked Network Seeks European Drone Operators and Urges Sabotage
A report published by the Cyfluence Research Center (CRC) states that a Russia-linked cross-platform cluster it tracks as UNK_SmoothOperator, which CRC began mapping in August 2026 and which appears to comprise more than 30 active core accounts and channels on X, Telegram, TikTok, Facebook, YouTube and Instagram, has combined pro-Russian messaging with overt attempts to recruit European citizens as drone operators for the Russian Armed Forces. On May 28th, 2026, 'David Z', an X persona CRC describes as a central node of the network, posted a recruitment appeal for the 'St. George' volunteer brigade offering a monthly salary of at least 220,000 rubles (about EUR 2,300) and the opportunity to obtain Russian citizenship, followed on 29 July by an appeal to 'determined Western citizens' with an emphasis on 'Irish patriots'.
Network nodes also amplified explicit calls for action against NATO targets, including sabotage of facilities supporting Ukraine: on April 15th, 2026, an X account that CRC associates with a British national who joined the Russian Armed Forces posted a 'call to action' targeting Ukrainian military equipment in the UK and Ireland that received more than 199,000 views, and pointed in a comment to a post by Dmitry Medvedev listing EU-based companies that supply components. CRC says behavioural patterns suggest a single actor or group manages the accounts, most of whose X accounts seem to connect from Ireland, warns that Europeans trained this way could potentially be deployed for hybrid attacks inside the EU or UK, and argues that counter-FIMI frameworks focused primarily on information manipulation are insufficient on their own.
Source: Cyfluence Research Center. CRC Spotlight: Propaganda, Recruitment, and Sabotage — Influence Networks as Emerging Hybrid Threats. [online] Available at: https://www.cyfluence-research.org/post/crc-spotlight-propaganda-recruitment-and-sabotage-influence-networks-as-emerging-hybrid-threats
[Appendix - Frameworks to Counter Disinformation]
UK Election Watchdog Develops AI Tool to Flag Deepfakes Quickly
An article published by Tech Policy Press states that the UK Electoral Commission is developing an AI-powered system designed to notify candidates of significant deepfakes, ideally within three hours of their posting on social media, according to chief executive Vijay Rangarajan. In a pilot during the May 2026 local elections, the system analysed about three million posts and flagged 147 items, of which human reviewers judged ten 'really quite misleading' and referred them to social media companies or the police.
Rangarajan described the approach as 'perpetrator neutral', focused on alerting candidates fast enough to respond before content goes viral rather than on prosecutions that would usually conclude long after an election. He also said the Commission faces 30,000 to 40,000 cyberattacks a month, sometimes peaking at 10,000 a day, and that it believes it has detected signs of 'algorithmic bias' on social media at election time, a problem for which he said the UK has no legislative solution.
Source: Tech Policy Press. UK Election Watchdog Develops AI System to Flag Deepfakes Within 3 Hours. [online] Available at: https://www.techpolicy.press/uk-election-watchdog-develops-ai-system-to-flag-deepfakes-within-3-hours/
Austria Plans Center of Excellence Against Foreign Information Manipulation
An article published by Vindobona.org states that Austrian Foreign Minister Beate Meinl-Reisinger announced in the Bundesrat the establishment of a 'Center of Excellence Against Foreign Information Manipulation', set to launch in spring 2027 with an international conference in Vienna and modelled on the European Centre of Excellence for Countering Hybrid Threats in Helsinki. Meinl-Reisinger said Russia's war against Ukraine has long since reached Central Europe and that a 'shadow war' is directly affecting Europe and Austria, citing past cyberattacks by Russian groups on the Ministry of Foreign Affairs and the Carinthian provincial administration, and said the government would strengthen media literacy and protect state institutions.
The plan drew a mixed reception: Freedom Party (FPÖ) lawmakers stressed permanent neutrality and warned of a potential 'censorship tool against dissenters', while NEOS and the Greens welcomed the course. Vindobona notes that critics fear a state-run centre could dictate what counts as disinformation and draw Austria closer to NATO-linked structures, while Vindobona adds that such a centre would give Austria access to threat assessments shared through the Hybrid CoE's network of 35 partner countries.
Source: Vindobona. Austria Plans New Center of Excellence to Combat Foreign Disinformation. [online] Available at: https://www.vindobona.org/article/austria-plans-new-center-of-excellence-to-combat-foreign-disinformation
AI-Generated Pre-Bunking Reduces Belief in False Election Claims
A study published by Washington University in St. Louis states that a study in Royal Society Open Science involving 4,293 registered U.S. voters found that AI-generated 'pre-bunking' articles significantly reduced belief in false election claims, even after participants read persuasive articles promoting those claims. Pre-bunking reduced immediate confidence in the false claim by 0.55 points on a 10-point scale, an effect that declined to a still statistically significant 0.17 points one week later.
Human experts first refined the prompt used to generate the articles, supplying the AI only with election-security information from the Cybersecurity and Infrastructure Security Agency, after which AI-generated interventions were no less effective than those given additional human review, which co-author Betsy Sinclair says could let defences operate 'at machine speed'. The researchers found no backlash in any partisan group but caution that the study tested only five rumours, measured self-reported beliefs rather than behaviour, and that the boost in confidence in election administration did not last a full week.
Source: Washington University in St. Louis. Using AI to Fight Election Misinformation Works, WashU Research Finds. [online] Published October 2026. Available at: https://source.washu.edu/2026/10/using-ai-to-fight-election-misinformation-works-washu-research-finds/
[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.
