top of page
CRC logo (2).png

Cyber based influence campaigns 19th – 25th October 2026 Report

Writer: CRC
CRC
11 minutes ago
28 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 19th to the 25th 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

Cyflunce Attack





[ Report Highlights]


• Graphika assessed with high confidence that Russia-aligned Storm-1516 produced very likely deepfake videos of a supposed former student accusing Senate candidate James Talarico of sexual assault.

• Senior Democrats Jim Himes and Mark Warner said Russia is attempting to harm the Democratic Party and erode voter confidence ahead of the midterm elections.

• Alliance4Europe attributed with high confidence to Russia's Doppelganger operation a bot network that turned French grievances against Ukraine before the October riots.

• Hacked correspondence obtained by The Insider shows Sergey Kiriyenko initiated the Doppelgänger fake-news operation, which Ilya Gambashidze's Agency for Social Design implemented.

• NewsGuard found that 176 of 189 X posts advancing false midterm election claims carried no Community Note, while four false claims advanced by Elon Musk carried none.

• Taiwan's security bureau said disputed or false online information spread by China rose about 48% to more than 2.5 million items in nine months.

• Pro-Ukrainian hacking group Black Spark claimed an attack that showed anti-Putin messages to students on Russia's school assessment platform, forcing assessments to be suspended.

• A Brennan Center retest of six AI tools found one company improved election safeguards, two made convincing fakes easier and the rest showed little change.

[ Report Summary]

• NewsGuard found that 176 of 189 X posts advancing false midterm election claims carried no Community Note, while four false claims advanced by Elon Musk carried none.

• TikTok removed examples flagged by Futurism of AI-manipulated arrest footage, discovered by Copyleaks, that falsely showed police officers kissing and groping arrestees.

• Graphika assessed with high confidence that Russia-aligned Storm-1516 produced very likely deepfake videos of a supposed former student accusing Senate candidate James Talarico of sexual assault.

• Two senior Democratic lawmakers said Russia is attempting to harm the Democratic Party and erode voter confidence ahead of the midterm elections.

• Alliance4Europe found no evidence that foreign actors started France's October protests but attributed them with high confidence to Russia's Doppelganger, a bot network that turned grievances against Ukraine.

• Hacked correspondence obtained by The Insider shows Sergey Kiriyenko initiated Russia's Doppelgänger fake-news operation, which Ilya Gambashidze's Agency for Social Design implemented.

• Jamestown reports that Russian forces are using AI-generated and manipulated videos to fabricate territorial gains in Ukraine, including around Kupiansk.

• DisinfoWatch identified 23 X posts that falsely portrayed Zelenskyy's Fox News ceasefire remarks as a Ukrainian defeat, together recording 17.89 million views.

• Taiwan's security bureau said disputed or false online information spread by China rose about 48% to more than 2.5 million items in nine months.

• The Taiwan Economic Democracy Union identified at least 28 Taiwanese entertainers who called the PRC their motherland on China's National Day, many reposting CCTV propaganda.

• IranWire documents how IRGC-affiliated outlets, state media, and Iranian officials spread false claims about France's protests, including a fake internet shutdown and a miscaptioned video.

• Ukraine's embassy in Iraq rejected as deliberate disinformation a Tasnim report claiming Ukraine had deployed a drone unit and thousands of drones in Iraq's Kurdistan Region.

• FakeReporter co-founder Achiya Schatz told the Reuters Institute his team faces a sharp increase in influence operations and manipulation campaigns ahead of Israel's October 27th election.

• Pro-Ukrainian group Black Spark claimed a cyberattack that disrupted Russia's nationwide school assessments and showed students messages urging them to join an anti-Putin movement.

• A Brennan Center retest of six AI tools found one company improved election safeguards, two made convincing fakes easier, and the rest showed little change.

• Check Point Research warns that AI-integrated antidetect browsers such as Kameleo, AdsPower and MultiLogin let a single operator run thousands of accounts while evading platform detection.

• Mississippi Attorney General Lynn Fitch said Senate candidate Scott Colom's AI ads appear to violate the state's deepfake law, and Colom said he would delete them.

• Spain's People's Party released an AI-manipulated campaign video depicting the prime minister's wife and Socialist figures appearing to admit wrongdoing, prompting outrage and a legal challenge.

• ISD's assessment of major platforms ahead of the US midterms found that transparency has not kept pace with AI integration as human oversight is scaled back.

• EU regulators are examining possible foreign online interference and disinformation around France's riots, including how cooperative social media platforms have been.

• NewsGuard identified three viral false claims, including AI-generated Eiffel Tower videos, apparently aimed at blaming France's riots on pro-Palestinian and Islamic militants.

• A study of 19 fact-checkers in Cameroon, Kenya, Nigeria and South Africa found that the political environment shapes both what they check and how they see their role.

• Graphika reviewed 100 TikTok accounts signalling at least support for Tren de Aragua and found 26 posting violent or threatening content, including posts identifying individuals as targets.

• DisinfoWatch found unsupported claims of a pipeline explosion, a higher death toll, and government cover-up reached large X audiences after a deadly fire at CFB Wainwright.

• Full Fact found Google's public SynthID Detector flagged five apparently genuine images shared by politicians as made or edited with AI, prompting Google to fast-track an update.

• Japan's internal affairs ministry released draft guidelines urging social media operators to curb election disinformation, including by suspending monetization and flagging AI-generated content.

• A draft from the European Parliament's Socialists and Democrats group calls for banning harmful, non-consensual or deceptive deepfakes, requiring platforms and AI providers to prevent their spread.

[State Actors]



X

X's Community Notes Missed 93 Percent of False Midterm Posts

An analysis published by NewsGuard's Reality Check states that X's crowdsourced fact-checking program, Community Notes, failed to flag 93.1 percent of the X posts advancing false claims about the U.S. midterm elections that NewsGuard identified. On Oct. 19, NewsGuard found 189 X posts advancing false midterm claims from its False Claims Fingerprints database; 176 of them carried no Community Note, and the unflagged posts advanced 23 of the 27 false claims tested, several of which had spread on the platform for months.

Examples include an unflagged March 2026 post presenting a court-ordered extension of voting hours for Democrats in Dallas County, Texas, as evidence of voter fraud, and NewsGuard stresses that the unchecked claims came from both conservative and liberal users. Community Notes flagged only four of the 27 falsehoods tested, and NewsGuard found that Elon Musk, who has 241.5 million followers, advanced four false midterm claims, none of which included Community Notes. X's press team did not respond to NewsGuard's request for comment.


Source: NewsGuard Reality Check. X’s Community Notes Flunk the Midterms. [online] Available at: https://www.newsguardrealitycheck.com/p/xs-community-notes-flunk-the-midterms 


Tiktok

TikTok Pulls Flagged AI Deepfakes of Police Groping Arrestees

An article published by Futurism states that engagement farmers on TikTok are using AI to manipulate real arrest footage so that police officers appear to kiss and grope people being taken into custody. The deepfakes, which Copyleaks discovered, are based on actual arrest footage that appears to have been originally filmed for reality TV shows such as TLC's “D.U.I.”, and are spliced in front of longer arrest videos, a tactic Futurism says suggests they were made to keep scrolling users watching through to the end.

Some of the clips disclose their use of AI and carry a disclaimer claiming they are for “entertainment use only with no harmful implication”, which Futurism says rings hollow because sexualized deepfakes of real people made without their consent are implicitly harmful. Copyleaks co-founder and CEO Alon Yamin said the videos “feature real people, both police officers and citizens, who are being depicted doing things that never happened”. After Futurism reached out, TikTok removed the examples Futurism had specifically flagged and pointed to its guidelines on clearly labeling realistic AI content.


Source: Futurism. TikTok AI Deepfakes Show Cops Groping Women. [online] Available at: https://futurism.com/artificial-intelligence/tiktok-ai-deepfakes-cops-groping-women


Russia

Storm-1516 Likely Behind Apparent Deepfake Accusing Talarico of Assault

An analysis published by Graphika states that a newly created X account seeded two very likely deepfake videos on October 20th that depicted a former student of Texas state Rep. and Democratic U.S. Senate candidate James Talarico accusing him of sexual assault. Based on dissemination tactics and content characteristics, Graphika assesses with high confidence that the Russia-aligned influence operation Storm-1516 produced the video.


The X account, created in September 2026 under the name Zyon Lewis, is very likely impersonating a Facebook user of the same name from Converse, Texas, who attended Rhodes Middle School, where Talarico taught, and who repeatedly expressed pride in having him as a teacher. Graphika points to unnatural or unsynchronized facial movements, a lack of body movement, and AI analysis tools as indicators of AI generation, and says known Storm-1516 amplifiers such as Liberal Tear Creator and Garbage Human, along with U.S. conservative accounts, shared the videos. In contrast, one QAnon Facebook account shared a link to them.


Source: Graphika. Storm-1516 Produces Very Likely Deepfake Videos of Student Accusing Democratic Senate Candidate. [online] Available at: https://www.graphika.com/insights/storm-1516-produces-very-likely-deepfake-videos-of-student-accusing-democratic-senate-candidate


Russia Seeks to Undermine Democrats Before Midterms

An article published by The Record states that the Russian government is attempting to harm the Democratic Party and erode voter confidence ahead of the midterm elections, according to Rep. Jim Himes, the top Democrat on the House Intelligence Committee, and Sen. Mark Warner, vice chair of the Senate Intelligence Committee. Himes said a classified U.S. intelligence community assessment shared with Congress in recent days also found China again engaging in targeted influence efforts, as are Iran and North Korea, and that “the Russians are kind of throwing everything out there, chaos, don't trust your institutions, voting machines”.


Warner said that “every election cycle since 2016, Russia has spread disinformation” to undermine Democrats, but both lawmakers hedged on how explicit the Kremlin's actions have been, suggesting that U.S. support for Ukraine is being used as a wedge issue. They stressed that the latest update showed no country pursuing efforts to hack into election infrastructure or tamper directly with votes, with Himes describing instead “a pretty robust influence environment”, and said they are assuming President Donald Trump will not accept results that are not in his favor and have wargamed a variety of scenarios.


Source: The Record. Democrat Lawmakers Say Russia Is Aiming to Undermine Party. [online] Available at: https://therecord.media/democrat-lawmakers-say-russia-is-aiming-to-undermine-party


Doppelganger-Linked Bots Turned French Grievances Against Ukraine

A report published by Alliance4Europe states that, while it found no evidence that a foreign actor was behind the origin of France's October 2026 protests, foreign actors worked actively to amplify and inflame public anger both before and after them. In the six weeks before October 17th, a bot network that Alliance4Europe attributes with high confidence to Russia's Doppelganger operation posted 639 coordinated messages in French, German, English and Polish, seen roughly 7.4 million times, which turned real grievances against Ukraine and mainly targeted centrist, democratic-left, pro-European and pro-Ukraine politicians.


The study, covering 547,471 posts on X, Telegram, TikTok, YouTube, Facebook and news sites in nine languages, found that once the riots broke out, the story was distorted abroad: roughly one French-language post in thirty-three directly blamed migrants or Islam, against nearly one in five in English. It also found that the Russian Pravda network published 1,018 articles across twenty-eight sites in nine languages, and notes that researchers argue the network is also designed to reach the data used to train AI models, a tactic they call “LLM grooming”.


Source: Alliance4Europe. EU Unrest Report · How France’s October 2026 Protests Were Manipulated, Then Retold Abroad. [online] Available at: https://alliance4europe.eu/report/eu-unrest-report-%c2%b7-how-frances-october-2026-protests-were-manipulated-then-retold-abroad


Hacked Files Expose Kremlin Team Behind Doppelgänger Fakes

An investigation published by The Insider states that hacked correspondence between Russian Presidential Administration staff and the Agency for Social Design (ASD) shows that Sergey Kiriyenko initiated the Doppelgänger fake-news operation, with Alexander Kharichev and Sofia Zakharova as its key Kremlin overseers and Ilya Gambashidze's agency responsible for its implementation. The team, set up in 2022 and tasked with “natively promoting Russian narratives in unfriendly countries,” fabricates stories in multiple languages and passes them off as Western media reports, while designers create deepfakes and strategists use AI to identify each country's most sensitive issues.


Reports compiled for the Presidential Administration collected reposts of the “Zelensky's yacht” story that Republican figures including JD Vance cited in December 2023, alongside material on the November 2023 Stars of David graffiti campaign in Paris, while AI prompts saved in April 2024 asked for comments written as a 38-year-old Frenchman, Polish woman or Israeli woman. The Insider concludes that the team has so far had more failures than successes, and notes that Gambashidze's claim of influencing opinion without being branded Russian propaganda sits uneasily with the sanctions imposed on him and some members of his team.


Source: The Insider. Putin’s “Doppelgängers”: The Insider gets inside the operations of the Kremlin's foreign disinformation network. [online] Available at: https://theins.press/en/inv/298359


Russia Uses AI-Faked Videos to Claim Battlefield Gains

An analysis published by The Jamestown Foundation states that Russian forces are using artificial intelligence-generated and manipulated videos to fabricate territorial gains in Ukraine, including around Kupiansk this month, supplementing traditional flag-raising operations to create the appearance of battlefield success for Russian and foreign audiences. On October 2nd, a Russian military blogger posted a Telegram video purporting to show the capture of part of Kupiansk-Vuzlovyi, but the terrain, especially the railway station building, does not match the current condition of the city's heavily damaged buildings, and Ukrainian experts say other visual anomalies also indicate the use of AI.


The analysis argues that false battlefield reporting can backfire by distorting Russian commanders' understanding of the war, pointing to a widespread practice of lying in reports to higher command that even Russian military bloggers have begun to criticize. It contrasts this with Ukraine's “Vivaldi” counteroffensive in the Lyman sector, where rigorous control over leaks, delayed disclosures, and the targeted destruction of Russian reconnaissance drones created an information vacuum that reduced Russia's operational intelligence and concludes that Russia's use of AI-generated video in military operations may quickly undermine trust in its information.


Source: Jamestown Foundation. Fabricated Russian Videos Backfire. [online] Published 21 October 2026. Available at: https://jamestown.org/fabricated-russian-videos-backfire/


X Posts Twisted Zelenskyy Interview Into False Defeat Story

An analysis published by DisinfoWatch states that within hours of a Fox News interview with Ukrainian President Volodymyr Zelenskyy on October 21st, prominent X accounts manipulated it to falsely claim that Ukraine had “lost the war” and that Zelenskyy had accepted defeat and agreed to surrender a quarter of his country to Russia. In fact, Zelenskyy said a freeze along current front lines could stop the killing, followed by talks and security guarantees, and rejected Russian pressure to abandon the rest of Donbas, while the “75% or more” figure came from the host's question, not from a Ukrainian concession.


Of 247 candidate X posts collected between October 21st and 22nd, DisinfoWatch classified 23 from a reviewed priority set as falsely portraying the interview as a Russian victory, Ukrainian defeat or permanent territorial concession, together recording 17.89 million views, and singled out American commentator Benny Johnson's video calling Zelenskyy “a defeated man.” It says the false victory story could serve both Kremlin messaging and MAGA-aligned commentators, but stresses that there is no evidence in its collection that the accounts acted under Russian direction or as part of a coordinated operation.


Source: DisinfoWatch. “Ukraine Has Not Lost, Russia Has Not Won”: How Zelenskyy’s Fox Interview Was Distorted. [online] Available at: https://disinfowatch.org/ukraine-has-not-lost-russia-has-not-won-how-zelenskyys-fox-interview-was-distorted/


China

Taiwan Reports Roughly 48% Rise in Chinese Disinformation Before Polls

An article published by Reuters states that Taiwan's National Security Bureau, in a report to lawmakers this week, said disputed or false online information spread by China and targeting Taiwan rose by about 48% in the first nine months of 2026 to more than 2.5 million items, from more than 1.69 million a year earlier, accusing Beijing of stepping up operations ahead of local elections. The bureau said the China-originated content focused on energy, food security and defence policy, as well as the platforms and images of specific candidates, in an effort to influence November's elections for city mayors and county magistrates.


According to the report, Beijing often created narratives through official media outlets before amplifying them through “online troll armies” and public relations firms, using tactics that included content farms, AI-generated videos, fake accounts and coordinated comment flooding, and “China has expanded the use of AI technology to manipulate information, seeking to influence voter support and election results”. The bureau said the messages were mainly distributed on social media platforms including Facebook, Threads, YouTube and online forums, and its data showed Chinese cyberattacks on Taiwanese government agencies averaging 2.58 million a day over the same period. China's Taiwan Affairs Office did not immediately respond to a request for comment, and Beijing has previously denied interfering in Taiwanese elections or spreading disinformation.


Source: Reuters. Taiwan Security Agency Flags Rise in Chinese Disinformation Ahead of Local Elections. [online] Published 21 October 2026. Available at: https://www.reuters.com/business/media-telecom/taiwan-security-agency-flags-rise-chinese-disinformation-ahead-local-elections-2026-10-21/


Taiwan Group Names at Least 28 Entertainers Echoing PRC Propaganda

An article published by Radio Taiwan International (RTI) states that the Taiwan Economic Democracy Union (EDU) said it identified at least 28 Taiwanese entertainers who publicly stated the PRC was their motherland on China's National Day on October 1, by reposting scripted propaganda from Chinese state media under their personal accounts or studio names, appearing in China Central Television's National Day programming or attending official receptions. According to EDU researcher Huang Cheng-han, 18 entertainers, including Peter Ho, Jiro Wang and Chen Linong, reposted CCTV's National Day posts on their personal Weibo accounts.


RTI reports that beyond reposting, Huang An and Ken Chu released congratulatory videos and actors Michelle Chen and Tim Huang took part in CCTV's National Day special, while Huang Cheng-han said Tim Huang also posted his invitation to the National Day reception at the Great Hall of the People and argued that such conduct amounted to cooperating with propaganda from an adversarial country. EDU think tank coordinator Lai Chung-chiang said the union respects entertainers' freedom to choose how to identify, but condemned retaining Taiwanese citizenship while cooperating with Chinese political propaganda and urged them to relinquish it before publicly supporting China.


Source: Radio Taiwan International (RTI). Taiwan Economic Democracy Union identifies 28 Taiwanese entertainers who participated in PRC propaganda campaign. [online] Available at: https://www.rti.org.tw/en/news?uid=3&pid=238697


Iran

Iranian Outlets Pushed Fake French Internet Shutdown Claim

An article published by IranWire states that France's protests are being met in Iran with distortions of reality and fabrications by Islamic Republic officials and the regime's propaganda apparatus. The IRGC and Iranian state media claimed that internet access in France had been shut down over student protests, a false report that originated from a high-traffic X account known for spreading misinformation and was repeated by Mohammad-Amin Aghamiri, Secretary of the Supreme Council of Cyberspace, while IRGC-affiliated outlets including Hamshahri and Tasnim published footage from a school in Florida as “riot police inside French schools”.


IranWire adds that state-aligned media figure Bashir Biazar told the Student News Network that some French students had died or entered comas as a result of police actions and that 36 percent of France's population is Muslim, claims it says are false, citing no recorded deaths or comas and an official 2020 survey in which about 10 percent of 18 to 59-year-olds identified as Muslim. It also notes that France summoned the Islamic Republic's ambassador on October 7th in protest against the “disinformation campaign by Iranian officials concerning the French student movement”.


Source: IranWire. Iranian State Media Spreads False Claims on French Unrest. [online] Available at: https://iranwire.com/en/features/159104-iranian-state-media-spreads-false-claims-on-french-unrest/


Ukraine Rejects Tasnim Claim of Drone Unit in Kurdistan

An article published by Kurdistan24 states that the Ukrainian Embassy in Iraq issued an official statement on Wednesday rejecting reports published on Tuesday by Iran's Tasnim News Agency, which cited an unidentified security source to claim that Ukraine had established a drone unit and deployed thousands of drones in the Kurdistan Region, describing the allegations as completely baseless and a form of deliberate disinformation. The embassy said Ukraine has not deployed any type of drone, drone unit, or related personnel in the Kurdistan Region or anywhere else in Iraq, and called on media outlets to refrain from publishing such unverified allegations.


Kurdistan24 notes that the Tasnim report was published the same day Ukrainian Ambassador Ivan Dovhanych was visiting Erbil for meetings with Kurdish political leaders and the Ukrainian community, and describes Tasnim, Iran's semi-official news outlet with close ties to the IRGC, as an active source of information operations targeting American-aligned actors in the region throughout the US-Iran conflict. The embassy also reaffirmed Ukraine's full respect for Iraqi sovereignty and its friendly relations with both the Iraqi federal government and the Kurdistan Regional Government.


Source: Kurdistan 24. Ukrainian Embassy In Iraq Rejects Iranian Media Claims Of Drone Deployment In Kurdistan Region. [online] Available at: https://www.kurdistan24.net/en/story/947566


Israeli Watchdog Tracks Surge in Election Influence Operations

An article published by the Reuters Institute for the Study of Journalism states that FakeReporter, an Israeli non-profit watchdog investigating malicious online activity, is grappling with a sharp increase in influence operations and coordinated online manipulation campaigns ahead of Israel's October 27th election, according to its co-founder Achiya Schatz. Schatz identifies three main conspiracy theories spreading online: claims of “treason from within” by traitors in the state apparatus who allegedly cooperated with Hamas on October 7th, claims that security chiefs refused to let Netanyahu know about the attacks, and a wave of fake news targeting Israel's Arab minority.


The article says FakeReporter recently uncovered a bot network operating through AI systems that campaigned for Itamar Ben Gvir against rivals within the Netanyahu camp, as well as a Likud SMS campaign against Avigdor Lieberman. On foreign interference, Schatz says “Iranians are always top players in the disinformation field”, describes an Iranian-linked account posing as an Israeli whose inflammatory posts were amplified by Netanyahu supporters, and warns that AI is increasingly powering micro-campaigns targeting specific audiences.


Source: Reuters Institute for the Study of Journalism. Meet the Israeli Watchdog Tracking Conspiracy Theories, AI Bots and Online Manipulation Ahead of the Election. [online] Available at: https://reutersinstitute.politics.ox.ac.uk/news/meet-israeli-watchdog-tracking-conspiracy-theories-ai-bots-and-online-manipulation-ahead


Cyfluence Attack

Hackers Push Anti-Putin Messages into Russian School Assessments

An article published by The Record states that Russia's education watchdog Rosobrnadzor says it has restored its online systems after a cyberattack on the previous Friday forced it to suspend assessments at schools and vocational education institutions across the country, with students completing questionnaires on its Edutest platform shown anti-government messages instead of questions, according to Russian officials. The Record reports that the messages urged students to join an underground movement against President Vladimir Putin, and a pro-Ukrainian hacking group calling itself Black Spark, which describes itself as an underground resistance movement operating inside Russia, claimed responsibility for the disruption.


The disruption coincided with a major stage of the assessment program, during which 9th-grade students in 29 Russian regions were scheduled to complete questionnaires, according to local media, and pro-Kremlin activist Ekaterina Mizulina said she had received reports from students across the Moscow region who encountered the messages. Rosobrnadzor did not disclose how the attackers gained access or whether any information was stolen, while Black Spark said schoolchildren had welcomed the disruption, denied recruiting minors, and encouraged adults to join its campaign against the Russian government.


Source: The Record. Anti-Putin Messages Disrupt Russian Education Platform. [online] Available at: https://therecord.media/anti-putin-messages-disrupt-russian-education-platform


[AI Related Articles]


Brennan Retest Finds Uneven AI Safeguards Against Election Fakes

An analysis published by the Brennan Center for Justice states that, rerunning weeks before the midterms the prompts it used earlier this year to test six popular AI tools, it found that one company significantly improved its protections against the creation of election disinformation, two others made it easier to create convincing fakes and the rest showed little change. OpenAI's ChatGPT made the most progress, now steering potentially harmful election prompts into a nonphotorealistic style, while Google's Gemini, using the Nano Banana 2.1 image model, now produces more realistic images, including some that look as if they were taken on a phone, and has removed the visible watermark that appeared on its images in the first round.


Meta AI, which in April refused to create a realistic photo of a draft executive order restricting mail voting, produced the image in October and appears to have stopped applying visible watermarks, while Runway and xAI's Grok again produced realistic images for nearly all prompts. The Brennan Center warns that open-weight models such as Black Forest Labs' Flux 3 Image Klein, which refused fewer harmful prompts than the company's proprietary Flux 3 Image, can be downloaded and stripped of their safeguards, and concludes that most major AI companies are taking minimal measures, arguing that legislators, regulators and state attorneys general must create rules and liability to limit misuse.


Source: Brennan Center for Justice. Major AI Models Are More Willing Than Ever to Create Fake Election Images. [online] Available at: https://www.brennancenter.org/our-work/analysis-opinion/major-ai-models-are-more-willing-ever-create-fake-election-images


Antidetect Browsers Let One Operator Run Thousands of Hidden Accounts

A report published by Check Point Research states that antidetect browsers obscure their users' digital identities, enabling a single operator to silently manage thousands of isolated accounts across platforms including Facebook, TikTok, Instagram and YouTube while evading the detection systems that look for scams and coordinated inauthentic behavior. Focusing on Kameleo, AdsPower and MultiLogin, which it selected based on documented references across underground forums and criminal marketplaces, the report lists confirmed use cases including account takeovers, carding, fraudulent loan applications and the bulk creation of artificial accounts.


Check Point says the tools' threat potential has increased with the integration of AI; AdsPower advertises a Model Context Protocol server that lets AI-generated content be produced, injected into pre-warmed account profiles, and distributed across platforms in a fully automated sequence, with its RPA Marketplace templates as the delivery mechanism, creating what the report calls an end-to-end influence operation pipeline that places coordinated influence operation capacity within reach of any sufficiently motivated actor. It warns that platform detection systems may struggle to address this threat surface, particularly around elections, and recommends shifting detection from the account and behavioral layer to the network and infrastructure layer.


Source: Check Point Research. The Risk of AI-Powered Antidetect Browsers in Online Fraud and Influence Operations. [online] Available at: https://research.checkpoint.com/2026/the-risk-of-ai-powered-antidetect-browsers-in-online-fraud-and-influence-operations/


Mississippi AG Says Colom AI Ads Appear to Break Deepfake Law

An article published by The Mississippi Independent states that Attorney General Lynn Fitch notified Democratic U.S. Senate candidate Scott Colom and his campaign on October 23rd, 2026, that artificial intelligence ads the campaign posted to social media appear to violate the state's deepfake law and should be taken down, warning that her office may go to court for an injunction. Colom afterward told Mississippi Today that he would delete the ad postings because he did not want Fitch to launch litigation using taxpayer money.


The notice followed a complaint from state Republican Party chair Mike Hurst about a 19-second video posted on Facebook and Instagram on October 1st showing a woman made to resemble Hyde-Smith, called “Cindy” by a narrator, hurrying away from reporters, captioned “Based on a true story”, with no indication that AI was used. The 2024 statute bars circulating a digitally fabricated depiction of a person within 90 days of an election without consent and with intent to injure a candidate, influence the result or deter voting, but exempts satire or parody without defining either; Colom called the video “clearly” parody, and Mississippi law does not require political ads to disclose AI use.


Source: Mississippi Independent. AG Warns Colom to Pull AI Ads or face injunction, and he agrees to do so. [online] Available at: https://msindy.org/p/ag-warns-colom-to-pull-ai-ads-or


Spain's Opposition Deepfake Ad Ignites Debate Over AI Campaigning

An article published by The Guardian states that a political ad from Spain's conservative People's Party (PP), launched this week across its social media channels, used AI to manipulate the likenesses of people linked to Prime Minister Pedro Sánchez's Socialist party, appearing to show them admitting to acts such as soliciting prostitutes and taking kickbacks. The video, which included an AI-generated version of Sánchez's wife, Begoña Gómez, and had racked up nearly 2m views by Friday, noted only in tiny, nearly transparent font that it was made with AI, and digital transformation minister Óscar López said it was “telling people a colossal lie”.


Former Socialist Party member Leire Díez, whose likeness was used, filed a court complaint reportedly accusing the PP of defamation and digital identity theft, although a majority of nine legal experts consulted by El País said the video was not illegal because it was labelled as AI-generated and satirical in tone. Researcher José Javier Olivas called the video a “red line” and said it “qualifies for disinformation,” while the PP's Borja Sémper said it used digital tools to highlight the Socialists' hypocrisy, and the Guardian notes that a research group at Wesleyan University identified more than 200 AI-generated or enhanced political ads in the US, the vast majority tied to Republicans.


Source: The Guardian. Opposition Deepfake Video Sparks Controversy Ahead of Spain Election. [online] Published 23 October 2026. Available at: https://www.theguardian.com/world/2026/oct/23/opposition-deepfake-video-spain-election


[General Reports]


ISD Finds Platform Transparency Lagging AI Integration Before Midterms

A report published by the Institute for Strategic Dialogue (ISD) states that the 2026 midterms are the first US election in which AI is built into every major social media platform, and that its central finding, after assessing Meta-owned Facebook and Instagram, X, YouTube, Snapchat and TikTok, is that transparency has not kept pace with AI integration. No platform has disclosed how its embedded chatbot will handle election questions, and ISD testing in June 2026 found that leading AI models, including X's Grok and Meta's Muse Spark, frequently gave inaccurate or incomplete answers to basic voter-information questions, with some responses potentially misleading or discouraging eligible voters from voting.


The report finds oversight shifting from people to crowd-sourced and automated systems: Meta ended its US third-party fact-checking partnership in January 2025, ISD's earlier review of the 25 most-voted Community Notes tied to the 2024 election found that only two of the 11 notes deemed accurate, necessary and consistent with X's standards were published, and the Financial Times reported in June 2026 that Meta had replaced around 50 percent of human review requests with LLMs. ISD also rates only TikTok as releasing findings on identified disinformation networks on an ongoing, monthly-or-more basis during the election cycle, and notes that none of the platforms responded to its requests for comment on the assessment.


Source: Institute for Strategic Dialogue (ISD). US Election Platform Preparedness 2026. [online] Available at: https://www.isdglobal.org/publication/us-election-platform-preparedness-2026/


EU Examines Possible Foreign Interference Around French Riots

An article published by the Financial Times states that EU regulators are examining possible foreign online interference and disinformation campaigns around the French riots and similar outbursts of violence and looting in other European countries, with an EU official saying, “We want to know how co-operative the platforms are”. According to security officials, pro-Russian and far-right social media accounts linked to previous disinformation campaigns have spread violent videos in an attempt to sow more chaos, and Sciences Po researcher David Colon said “pro-Kremlin influencers” linked to the Russian campaign Storm-1516 have “amplified the events and spread several false narratives,” while stressing they “are not at the origin of the demonstration, nor have they had an impact on the course of the events”.


One senior European intelligence official said there was no concrete evidence that Russian agents directly stoked street violence, describing Russian disinformation efforts as mostly reactive and online while noting that French investigations were at an early stage, and another said foreign actors' role “should not be overestimated.” The FT identified several accounts posting violent content, including footage falsely claimed to have been recorded in France, such as an X video of protesters throwing fireworks, viewed almost 6mn times, that was in fact recorded in 2024 during protests in Tbilisi, Georgia, and the Commission said Brussels would step in if the risk became “systemic” and online platforms failed to co-operate.


Source: Financial Times. EU examines pro-Kremlin disinformation around French riots. [online] Available at: https://www.ft.com/content/d95609cd-1bb9-4995-937b-875277de3a42


NewsGuard Debunks Viral Fakes Apparently Blaming Riots on Islamists

A report published by NewsGuard's Reality Check states that as mass protests in France spiraled into rioting, arson and looting, far-right social media users pushed false claims apparently aimed at depicting the unrest as driven by pro-Palestinian immigrants and Islamist terrorists. NewsGuard identified three viral false claims, including AI-generated videos showing protesters unfurling a giant Palestinian flag from the top of the Eiffel Tower, shared in one X post viewed 3.7 million times, and of the tower engulfed in flames and toppling, in a post that garnered 11.6 million views, both of which AI-detection tools indicated were inauthentic or AI-generated.


A third claim, pushed by far-right and pro-Israel users, many of whom shared a CNews clip in which journalist Paul Amar said he had heard that all the stores owned by Jews on a street in Marseille were set on fire, alleged that Jewish shops had been singled out; NewsGuard calls it baseless, citing French reports that Muslim and Catholic store owners were also hit, denials from local authorities and Jewish leaders according to Libération, and a statement by France's Protection Service for the Jewish Community (SPCJ) that there is no evidence the attacks were targeted based on religious affiliation. NewsGuard notes that the protests began in late September over teacher shortages and deteriorating school facilities before widening into an anti-government movement, with approximately 2,500 people arrested over the weekend, according to the French Justice Ministry.


Source: NewsGuard Reality Check. French Unrest Inflamed by Misinformation. [online] Available at: https://www.newsguardrealitycheck.com/p/french-unrest-inflamed-by-misinformation


Political Systems Shape What African Fact-Checkers Check

An article published by The Conversation states that a study of fact-checkers in Cameroon, Kenya, Nigeria and South Africa, based on in-depth interviews with 19 professionals between January and August 2025, found that the political environment shapes not only what fact-checkers can do but what they believe their job is. Cameroonian fact-checkers, working under a government the authors describe as authoritarian in practice, said they mostly stayed away from politics out of fear and checked claims about fires, accidents, and celebrities instead, while Kenyan fact-checkers saw themselves as helping people take part in democracy and holding leaders accountable.


Senior Nigerian fact-checkers described serious pressure, one citing warnings “sometimes framed as 'friendly advice' but clearly meant to intimidate” once a claim touches politics or security and another the use of cybercrime and defamation laws to threaten legal action, while South African fact-checkers saw themselves as a complement to struggling newsrooms. Across all four countries, many are moving beyond true-or-false verdicts toward “prebunking”, and the authors call on governments to stop misusing the law to intimidate fact-checkers and on funders and technology companies to offer long-term support.


Source: The Conversation. What Gets Fact-Checked? How Politics Shapes the Answer in Four African Countries. [online] Available at: https://theconversation.com/what-gets-fact-checked-how-politics-shapes-the-answer-in-four-african-countries-293964 


Graphika Finds Violent Content Among Pro-Tren de Aragua TikTok Accounts

A report published by Graphika states that it identified a cluster of TikTok accounts displaying textual and visual characteristics associated with the Venezuelan criminal organization Tren de Aragua (TdA), such as train and pirate references, coded emojis and references to the group's leadership and places of origin, which signify at least support for TdA, if not direct association, although Graphika did not identify open-source indicators that definitively tie the accounts to TdA members. Reviewing a sample of 100 accounts between September 12th and 25th, 2026, it found that 26 posted violent or threatening content, including depictions of killings framed as TdA actions and posts identifying individuals as targets, which other accounts acknowledged in the comments.


The report also found recruitment posts offering unspecified jobs to audiences inside Venezuela and other countries, including the Venezuelan diaspora, that directed interested users to private messages and, in one case, a WhatsApp group, while 96 accounts across nine countries used both train and pirate references, only two of them with imagery or flag emojis suggesting they were in or from Chile, which Graphika says suggests the use of pirate references is becoming widespread among TdA supporters. The accounts attempted to evade moderation by masking weapons and violence with emojis and spreading identity markers across images and captions, but some operated openly, with a September 2026 TikTok search for “tren de aragua” returning multiple accounts promoting or supporting TdA; Graphika stresses that no single indicator should be read as evidence of affiliation on its own.


Source: Graphika. Tren-ding on TikTok. [online] Available at: https://www.graphika.com/reports/tren-ding-on-tiktok


Pipeline and Cover-Up Rumours Spread on X After Wainwright Fire

An analysis published by DisinfoWatch states that after four Canadian soldiers were killed and 39 injured when a fire swept through a training area at CFB Wainwright, Alberta, on October 17th, social media saw a rush of rumours, including claims of a gas pipeline explosion, a much higher death toll and suggestions that the federal government was hiding what really happened. Reviewing 2,917 X posts with about 32.2 million recorded views, it identified 61 posts repeating one or more of six disputed claims, with exaggerated casualty claims drawing the most recorded views and cover-up accusations the most replies.


DisinfoWatch notes that the Canadian Armed Forces' October 22 update said the fire began during routine training on the Automatic Grenade Launcher range and intensified after a change in wind, and that the Provost Marshal said there was no indication natural-gas infrastructure was involved. Russian state-controlled RT, whose four posts in the dataset drew approximately 606,000 combined views, and U.S. outlets such as The Gateway Pundit amplified parts of the story, but the report says this does not establish coordination or who originated the rumours, and that their impact so far appears marginal.


Source: DisinfoWatch. Wainwright Fire Sparks Online Conspiracies and Cover-Up Allegations. [online] Available at: https://disinfowatch.org/wainwright-fire-sparks-online-conspiracies-cover-up-allegations/


[Appendix - Frameworks to Counter Disinformation]


Google Updates SynthID Detector After Full Fact Finds Likely False Positives

An investigation published by Full Fact states that Google fast-tracked an update to the public SynthID Detector tool it launched earlier this month after a Full Fact investigation found it describing apparently real images as “made or edited” with AI. Full Fact found five examples of images shared by politicians in recent months, including Plaid Cymru's Niamh Salkeld and Labour MP Nadia Whittome, that the tool flagged even though those who posted them insist they are genuine and unaltered.


After Google rolled out the update on October 15th, the tool changed its verdict on the five images to “Unsure,” but Full Fact says questions remain, citing a September image posted by Sinn Féin MP Órfhlaith Begley that was still flagged although she said no AI was used. Describing the cases as likely false positives because it cannot completely rule out minor inadvertent AI edits, Full Fact says SynthID remains an invaluable tool but that it never relies on it alone to prove whether an image is fake or real.


Source: Full Fact. Google SynthID Detector Flagging Apparently Real Images. [online] Available at: https://fullfact.org/toolkit/google-synthid-detector-flagging-apparently-real-images/


Japan Drafts Social Media Guidelines Against Election Disinformation

An article published by Nippon.com (Jiji Press) states that Japan's internal affairs ministry on Friday released draft guidelines outlining measures for social media operators to mitigate the adverse effects of false or misleading information online on elections. The guidelines cite suspending monetization of posts that undermine electoral fairness, and detecting images and videos generated by artificial intelligence and displaying warnings about them, as examples of recommended measures.


The draft, compiled in response to the revised public offices election law and the revised information sharing platform law enacted in July, will be finalized as early as the end of this year so that the guidelines can be in place for the unified local elections in April next year. Jiji notes concerns about systems that let users earn money based on the number of views or video plays their posts receive, which have been blamed for contributing to the spread of radical content and false or misleading information during campaigns.



EU Socialists Push Ban on Harmful, Non-Consensual or Deceptive Deepfakes

An article published by POLITICO Europe states that a draft of the tech objectives of the European Parliament's center-left Socialists and Democrats group, seen by POLITICO, calls for a ban on harmful, non-consensual or deceptive deepfakes, a move that would require platforms and AI providers to prevent their spread rather than wait for victims to complain. German S&D MEP Tiemo Wölken warned that “the current trend towards different national solutions only creates legal confusion bad actors could exploit.”


The push comes as Denmark this month introduced a plan to amend its copyright law to bar sharing digital replicas without consent, and after a report that 50 MEPs have been targeted by sexualised deepfakes. Legal experts warn that copyright-style protections could let companies acquire control over people's AI-generated likenesses, and POLITICO notes that while the AI Act sets transparency requirements for deepfakes and the Digital Services Act tells platforms how to deal with illegal content, none of the existing rules create a general EU-wide right over one's digital likeness; a Commission consultation on copyright rules runs until November 3rd.


Source: POLITICO Europe. Socialists and Democrats Push for EU Law on Deepfakes. [online] Available at: https://www.politico.eu/article/socialists-and-democrats-push-eu-law-deepfake/


[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