Dear Disinfo update reader,

In this edition of the Disinfo Update, we look at another turbulent month in the information space. The European Commission published its Communication on the Democracy Shield on 12 November. EU DisinfoLab welcomes the Commission’s decision to prioritise the protection of democracy and the fight against foreign interference and disinformation. However, the Shield falls short of the new capabilities that were promised. It overlooks the deep structural problems shaping today’s information ecosystem, from the dominance of big tech platforms and their opaque recommender systems to an online advertising model that fuels clickbait, hyper-personalisation and fraud. Instead of obliging platforms to comply with the law, many proposed measures shift responsibility onto users or rely heavily on voluntary measures from actors who already fail to comply with existing obligations, while also repackaging several old or pre-existing initiatives as if they were new. Read more about the Commission’s Democracy Shield communication and our in-depth analysis on the EU DisinfoLab website. 

We also dive into foreign interference attempts, platform transparency issues, and new research tracking the evolution of influence operations across Europe. From pro-Russian DDoS attacks on Danish political actors to the exposure of coordinated Chinese espionage efforts on LinkedIn, the landscape continues to demonstrate how adaptable and persistent hostile actors have become. 

Across the tech ecosystem, platforms are introducing new measures– some promising, others imperfect. TikTok is testing invisible AI watermarking, X continues to generate controversy after its new transparency tool revealed foreign origins of prominent political influencers, and researchers warn about the inadvertent amplification of the pro-Russia Pravda network by mainstream outlets. Together, these developments underscore a growing disconnect between the scale of manipulation risks and the unevenness of our current safeguards. 

Get reading and enjoy!

Our Webinars

UPCOMING – REGISTER NOW!

  • 4 December: Starting Over: A new strategy for US information integrity | The United States government has chosen to unilaterally disarm in the global contest for truth and integrity, even as authoritarians and technology companies find increasing alignment. The abrupt shuttering of our information integrity instruments comes with significant risks, leaving Americans, and people worldwide increasingly exposed to an information ecosystem designed to manipulate the public and erode social trust. The global information space is closing; the U.S. integrity community must now critically assess our past efforts, confront difficult realities, and urgently chart a new, resilient strategy for the road ahead. Failure to do so would be a massive win for authoritarians and their repressive global agenda. Join Adam Fivenson as he explores what “starting over” should look like, the hard lessons that must be confronted, and the strategic choices that will define the future of democratic resilience.
  • 11 December: Are platforms curbing disinformation? Scientific, cross-platform evidence from Six VLOPs | How much mis- and disinformation do people actually see on major online platforms? Are repeat misinformers being structurally rewarded with extra reach and monetisation? There has been no scientific, cross-platform way to answer these questions… until now. This webinar will introduce the first results from SIMODS — Structural Indicators to Monitor Online Disinformation Scientifically, the first project to measure the prevalence of mis- and disinformation scientifically across platforms and languages. Emmanuel Vincent from Science Feedback will walk us through the methodology behind this groundbreaking effort and reveal several striking findings. He will also address a key concern for our community: the significant data-access barriers researchers still face and demonstrate how this type of rigorous, cross-platform measurement is crucial for building the Structural Indicators envisioned in the European Code of Practice on Disinformation.

PAST – WATCH THE RECORDINGS!

  • Command and Control: How ANO Dialog surveils the Russian info space for the Kremlin | Behind Russia’s polished state messaging lies a vast monitoring apparatus: ANO Dialog. As the nerve centre of the Kremlin’s information control, it quietly manages propaganda and online manipulation across thousands of social media channels. In this session, Serge Poliakoff (University of Amsterdam) uncovers how ANO Dialog works, why it represents a new model of state-controlled disinformation, and what its reach means for the information space today.
  • Weaponising the past: historical disinformation as a tool to legitimate aggressive politics | This session traces how history and archaeology have been manipulated to legitimise power, first by Nazi Germany and today by the Russian state. Chiara Torrisi, historical researcher, will show us how old mechanisms of distortion have been adapted to today’s information environment, turning the past into a political weapon.
  •  Are AI Detection Tools Effective? Introducing TRIED, a WITNESS Benchmark | How do we ensure AI detection tools truly serve the people who need them most—and strengthen the work of fact-checkers, journalists, and civil society groups? In this webinar, Zuzanna Wojciak presents TRIED: the Truly Innovative and Effective AI Detection Benchmark, a practical framework developed by WITNESS, to evaluate whether detection tools are genuinely useful and effective while simultaneously guiding AI developers and policymakers in designing and promoting inclusive, sustainable, and innovative detection solutions.

Disinfo news & updates 

🏛️Policy and regulation  

  • EU unveils Democracy Shield. The European Commission has unveiled its European Democracy Shield initiative, which is aimed at countering disinformation, fake news, and hybrid threats linked to foreign interference as Europe heads into key elections. The plan places new expectations on major online platforms like Google, Meta, and X, urging them to go beyond existing Digital Services Act rules by more actively detecting and labelling AI-generated or manipulated content. The plan also includes a voluntary influencer network to boost awareness of political campaigning rules, and it establishes a European Centre for Democratic Resilience to increase support for independent fact-checkers and local media to reinforce the integrity of public information. Critics, however, say the initiative “lacks teeth,” according to a DW article, noting that many of its core measures remain voluntary for member states wary of ceding sovereignty. The EU is also treading carefully to avoid friction with the US, particularly amid the Trump administration’s hostility toward new digital rules, by favouring voluntary cooperation from tech platforms over binding regulation. 
  • AI Regulation. The European Commission has proposed delaying key “high-risk” provisions of the EU AI Act until December 2027, following sustained pushback from Big Tech and concerns about competitiveness. The plan, part of a wider ‘Digital Omnibus’ to streamline digital rules, would also ease GDPR restrictions by allowing companies like Google, Meta and OpenAI to use Europeans’ personal data to train AI models. Brussels insists the move is “simplification, not deregulation,” but critics warn it marks a significant weakening of the EU’s landmark AI framework.  
  • Spanish court orders Meta to pay for GDPR violations. A Madrid commercial court has ordered Meta to pay 542 million euros to 87 Spanish media outlets after ruling that the company engaged in unfair competition. The case centred on Meta’s unlawful processing of vast amounts of personal data to deliver highly targeted ads without obtaining proper GDPR consent, giving the platform a dominant advantage over traditional publishers. The judgment marks a major precedent for the digital advertising market in Spain and across Europe, reinforcing that data-protection breaches can carry significant competition consequences for tech giants. 
  • Coimisiún na Meán’s Investigation into X. Coimisiún na Meán (CNAM), the Irish regulator, has launched a formal investigation into X under the EU’s Digital Services Act (DSA), focusing on potential breaches of Article 20. The inquiry follows concerns that X may not provide an effective internal system for users to appeal decisions on reported content, nor adequately inform them of their rights or the outcomes of their reports. As the EU’s Digital Services Coordinator for Ireland, Coimisiún na Meán is responsible for ensuring platforms uphold the DSA’s core principle that users must be able to contest content-moderation decisions. If X is found non-compliant, Coimisiún na Meán can impose major administrative penalties, including fines of up to 6% global turnover.

📰Information manipulation by media distortion

  • Investigation finds Fox News misled viewers on Portland protests. A ProPublica investigation reveals that Fox News aired a distorted narrative of escalating violence during Portland’s anti-ICE protests, a portrayal that preceded President Trump’s decision to deploy federal troops. The report shows how the network used outdated footage from the 2020 George Floyd protests while presenting it as events currently happening in 2025, pairing it with dramatic headlines such as “riots raging.” ProPublica’s review of video evidence and arrest records found no pattern of protestors initiating violence. Instead, further investigations revealed that the recent clashes largely involved federal officers using force without corresponding charges against demonstrators. The investigation also shows Fox News misrepresented the timing of local incidents to reinforce an image of a city in chaos. Overall, the findings suggest the network skewed its coverage to justify federal intervention.

💻Platform transparency

  • Grokipedia cites a Nazi forum. A new analysis by Indicator finds that Elon Musk’s AI-driven encyclopedia, Grokipedia, leans heavily on fringe and unreliable sources, far more than Wikipedia. Researchers discovered millions of citations from low-quality domains, including conspiracy outlet InfoWars and the white nationalist forum Stormfront, which Wikipedia rarely or never uses. While much of Grokipedia mirrors Wikipedia’s content, a troubling subset of entries includes large numbers of dubious citations, often on sensitive topics. The findings suggest an editorial direction that may elevate specific narratives rather than factual accuracy. Thus, contradicting Musk’s claim that the platform aims to deliver “the truth, the whole truth, and nothing but the truth.” 
  • TikTok now allows users to choose how much AI-generated content appears in their feeds. The platform is introducing an update that lets users adjust how much AI-generated content (AIGC) appears in their “For You” feed through the existing “Manage Topics” setting. The change comes as hyper-realistic AI content becomes more widespread across social platforms, following similar moves by Meta and OpenAI to integrate AI-driven feeds. To enhance transparency, TikTok is also experimenting with a new form of “invisible watermarking,” designed to offer stronger identification of AI-generated media than current approaches like C2PA’s Content Credentials. In addition, the company has launched a $2 million AI literacy fund to help users better understand AI safety, detection, and the evolving media landscape.
  • X’s new transparency tool reveals foreign origins of prominent “America first” accounts. A Daily Beast investigation shows that a new “About this Account” transparency feature on Elon Musk’s platform X has inadvertently exposed several well-known MAGA influencers as operating from abroad. Despite promising an “America first” agenda, many of these accounts were shown to be based in countries including Russia, India, and Nigeria. The revelations triggered a wave of commentary from Democratic and left-leaning influencers, who argued the findings validate long-standing concerns about foreign influence operations shaping US political discourse. Although some foreign operators may simply be chasing X’s engagement-based payouts, the disclosures raise deeper questions about the scale and impact of non-US actors in American online politics. While X’s new location feature can help show who is really behind an account, it is also being misused. According to a review by NewsGuard, some partisan accounts are now creating fake screenshots and fake profiles to wrongly accuse others of hiding that they are from another country.

🕵️Cyber threats and espionage for FIMI purposes

  • Cyberattacks for FIMI in Danish elections. Several Danish political parties and government agencies briefly saw their websites knocked offline following DDoS attacks claimed by the pro-Russian group NoName057(16). Authorities said such cyber attacks have become “part of the normal picture” in Denmark and noted an uptick in incidents in the day before the vote. Crucially, officials confirmed the attacks had no impact on the election itself, as all voting is conducted manually. The incidents reflect a wider pattern of pro-Russian hacking groups targeting elections across Europe, even as law enforcement works to dismantle their operational infrastructure.
  • Chinese espionage against British MEPs on LinkedIn. MI5 has issued an urgent alert to MPs after uncovering a sweeping Chinese espionage effort using LinkedIn to obtain sensitive political intelligence. According to the warning, China’s Ministry of State Security has been deploying fake profiles and civilian “head-hunters” to build long-term relationships with MPs, parliamentary staff, and policy experts. Security Minister Dan Jarvis said the UK “will not tolerate” such interference, noting it is part of a broader pattern that includes cyber-operations and earlier attempts to influence British democracy.     
  • Leak exposes scale of global Chinese hacking operations. A major data breach at Chinese cybersecurity firm KnownSec has exposed more than 12,000 confidential documents revealing the breadth of Beijing’s state-aligned hacking operations. The leaked files detail an extensive cyber arsenal– from Remote Access Trojans to hardware-based exploits. The documents also confirm the large-scale theft of sensitive data from dozens of countries, including telecommunications and immigration records, illustrating the global reach of China’s cyber-espionage network.

🤖AI Disinfo updates

  • AI (mis?)used for fact-checking. Two new investigations shed light on X’s rapid turn toward AI-driven fact-checking. The Columbia Journalism Review revealed that AI bots are now writing a growing share of Community Notes and, in some cases, generating inaccurate drafts that spread before verification, as seen in the mislabelled No Kings protest video. Another analysis by Indicator shows the scale of this shift: one AI bot, “Zesty Walnut Grackle,” has already become the platform’s top contributor, publishing more than 1,100 notes since September. Together, the reports highlight how AI is reshaping crowdsourced fact-checking on X, raising questions about accuracy, governance, and the future of platform-based verification.
  • Research on AI. People are more likely to believe false headlines when they are paired with highly realistic AI-generated images that appear to provide strong evidence, according to a new peer-reviewed study from the Harvard Kennedy School. The research suggests that photorealistic synthetic visuals increase susceptibility to mis-/disinformation and that directing users’ attention to image properties may help reduce belief in false claims.
  • Cyberattacks and cybersecurity. AI is increasingly shaping both cyberattacks and cyber-defence. According to the BBC, Anthropic says a Chinese state-linked group used its Claude model to automate parts of a cyber-espionage campaign, though Ars Technica and CyberScoop report the operation was far from “autonomous”, relying on human-built frameworks and frequent corrections. Amid rising concern over AI-enabled threats, Germany’s BSI has issued new guidance to help organisations defend large language models against “evasion attacks” such as jailbreaks, prompt injections, and other adversarial inputs, offering practical steps to harden LLM deployments.
  • AI for good. AI is increasingly being applied to strengthen information integrity, support verification, and help counter fast-moving falsehoods. Full Fact, Maldita.es, and the European Fact-Checking Standards Network have launched Prebunking at Scale, an AI-driven system that spots emerging disinformation narratives on short-form video platforms. Meanwhile, AI-driven tools are helping to detect AI-generated audio.  DW Innovation tested three of them (Hiya, Deepfake Total and DeepFake-O-Meter), finding that none can reliably identify AI-generated voices across languages and formats. Each tool has partial strengths but inconsistent results, reinforcing that audio-deepfake detection still needs to be paired with traditional verification.

Want to stay on top of the latest in AI and disinformation? Our AI Disinfo Hub has just been updated. Take a look!

🌱Clarity Clarity hub

  • “COP of Truth.” At COP30 in Belém, Brazil, countries formally committed for the first time to tackling climate disinformation through a new Declaration on Information Integrity. The declaration, according to EuroNews, has been signed by 12 countries: Belgium, Brazil, Canada, Chile, Denmark, Finland, France, Germany, the Netherlands, Spain, Sweden and Uruguay.
  • Who Tried to FLOP COP30? A group of politicians, right-wing outlets and verified influencers pushed the #FLOP30 disinformation campaign, an anti-COP campaign, amplified by figures like Donald Trump, Friedrich Merz, Aldo Rebelo, and several Brazilian lawmakers, according to Oii – Observatory for Information Integrity – Climate.
  • Greenwashing blitz before COP30. Recent investigations reveal massive coordinated campaigns ahead of the summit: DeSmog reveals that big Ag (big agriculture) hired 195 Brazilian influencers to flood Instagram with glossy, climate-free meat and agrochemical promotions, double last year’s total. Oil & gas giants ran an unprecedented wave of digital greenwashing, as shown by CAAD, with a 2,900% surge in Google Ads targeting Brazil in 2025, aiming to soften scrutiny on emissions, deforestation, and health impacts.
  • AI: Tool or Threat? AI became a central theme at COP30, praised for its potential in emissions tracking, climate adaptation, and Global South innovation, as published by The Conversation. But investigators also uncovered coordinated AI-generated disinformation campaigns, according to The Standard. This includes synthetic videos, cloned voices, and automated networks designed to polarise debate and undermine trust in climate negotiations.

Want to stay on top of the latest in climate change and disinformation? Our Climate Clarity Hub has just been updated. Take a look!

Brussels Corner

Thierry Breton on the Debate Over EU Tech Sovereignty

The Guardian this week published an opinion piece by former Commissioner Thierry Breton, setting out his vision of what “European technological sovereignty” should mean. The concept has become central to current policy debates, yet remains relatively vague.

Most notably, he delivers a sharp critique of the current Commission leadership’s approach to tech regulation. Breton warns against efforts to dilute cornerstone EU digital laws “through ‘omnibus’ bills or otherwise,” just months after they entered into force. Framing these moves as driven by transatlantic pressure, he urges Europe not to become “useful idiots” in attempts to unravel its own digital frameworks. Protecting the integrity of the EU’s digital rulebook, he argues, is essential not only for innovation but also for Europe’s geopolitical positioning.

He makes the same argument in a piece published by Handelsblatt. For a former commissioner, this is unusually strong language — highlighting just how sensitive and politically charged the debate around Europe’s tech sovereignty has become

First DSA Article 35(2) Report Published

The European Commission and the Board of Digital Services Coordinators have published the first Article 35(2) DSA report on the main systemic risks found on very large online platforms and search engines (VLOPs & VLOSEs) in the EU. The report covers the period from 17 February 2024 to 16 February 2025. The wording of article 35(2) is very flexible – allowing the inclusion of systemic risks identified by very large online platforms and search engines, but also information “from other sources,” offering the opportunity to identify and detail systemic risks in a broad and effective way. 

Article 35 requires big platforms to take reasonable and effective steps to reduce the systemic risks identified under Article 34. This can include changes to how their services are designed, how content is moderated, how algorithms or ads work, how they cooperate with trusted flaggers or other platforms, and how they protect minors or label AI-generated and manipulated media. The new report is a chance for the Board to look beyond the list in the law and identify a wider range of systemic risks emerging in the EU’s online environment. Informally, there seems to be a consensus in the Board and Commission that this first report was a learning experience and that future iterations will be more ambitious.

Reading & resources

  • ‘Romanian scenario’ in Czech elections. New research from the Institute of Strategic Dialogue (ISD) charts the rapid spread of the so-called “Romanian scenario” conspiracy theory in the run-up to Czechia’s 2025 parliamentary elections. Originating from claims that Romania’s 2024 elections were annulled due to alleged Russian interference, the narrative morphed into a broader accusation that the EU and Czech institutions were preparing to rig the upcoming Czech elections. ISD’s analysis shows that far-right and far-left politicians, alternative media outlets, and Russia-linked channels played key roles in amplifying the narrative, fueling public distrust in democratic institutions such as the Constitutional Court and the intelligence services. By examining activity on X, the study illustrates how conspiratorial framing can rapidly mutate, connect disparate domestic and international disputes, mobilise voters, and ultimately undermine confidence in electoral integrity.      
  • AI-generated photo guessing game. Can you recognise a real photo among AI-generated images? Test your skills in this fast-paced photo challenge by FaktaBaari. 
  • ISD Pravda network. Recent research from the Institute for Strategic Dialogue (ISD) shows that hundreds of webpages, including many English-language news and academic sites, frequently cite the pro-Russia Pravda network without acknowledging its role in a long-running Russian information operation. The study found that 81% of reviewed sites treated Pravda content as credible, unintentionally boosting the network’s visibility and helping its material surface more prominently in search results. To counter this inadvertent legitimisation, ISD has launched a free browser extension that alerts users whenever they land on a Pravda network webpage. 

This week’s recommended read

This edition’s recommended read is brought to you by EU Disinfo Lab’s very own Executive Director, Alexandre Alaphilippe. 

Sometimes we discover a book so essential that we wonder how we missed it for so long. This week, based on the advice of my colleague Gary, I finally took the time to read Timothy Snyder’s On Tyranny, which was released eight years ago, around the time when EU DisinfoLab was born. Given today’s political climate, its lessons feel even more urgent. On Tyranny is a short book that lays out 20 crucial lessons on how to resist authoritarianism.

Snyder describes with great clarity the mechanisms by which authoritarian forces take only the power that society concedes to them. As a consequence of this abandonment of responsibility, he explores how political submission to the agenda of disinformation agents and the lack of concrete action to defend institutions slowly push the rule of law toward the rule of the strongest.

To resist this slide, Snyder argues that each of us — in line with our own capacities — must act: to defend truth and facts, to engage in public debate, and to build strong networks of solidarity.

For me, reading these lessons has been motivating and has helped me reflect once again on why we remain engaged in civil society. In the current environment, we must be especially wary of the “politics of inevitability” and the “politics of eternity.” History is not predetermined; it depends on our actions to defend rights and hold accountable those who try to erode them.

The latest from EU Disinfo Lab

  • Mapping algorithmic amplification: lessons from Germany. We’re excited to share our latest report by our research manager, Maria Giovanna Sessa, and our senior researcher, Raquel Miguel Serrano, which explores how algorithmic amplification shapes content visibility across major platforms and why this process remains a “black box” for regulators, researchers, and the public. Drawing on four German cases involving TikTok, X, Amazon, and Meta, the report shows how legal action is beginning to push platforms toward greater accountability. It also highlights how amplification can create systemic risks by distorting information flows and influencing public debate.  
  • Updated factsheet: disinformation landscape in Bulgaria. We have just released an updated Bulgaria Disinformation Landscape Factsheet, expanding on our 2023 edition with the latest developments shaping the country’s information ecosystem. From ongoing political volatility and media-sector vulnerabilities, to the rise of AI-generated false content and the growing role of grassroots fact-checking, the factsheet provides a comprehensive look at how disinformation evolves in Bulgaria today. A huge thank you to Ruslana Margova and Patrick Gensing for their expertise and collaboration.   
  • Malicious semi-compliance. A new blog post by our policy and advocacy intern, Katrina Luize Asmane, and our senior policy expert Joe McNamee has just been published! The blog post examines how big tech platforms increasingly rely on “malicious semi-compliance” as a strategy to appear aligned with the DSA while avoiding meaningful compliance. Through three concrete cases, it shows how design choices and political ad bans shift the burden of enforcement onto regulators, researchers, and civil society.

Events & announcements  

  • 2 December: Meet the Future of AI: Support for the European Democracy Shield: policy, practice and partnerships. Join policymakers, researchers, and tech experts as they tackle the growing threat of online disinformation. This event explores how AI, cross-sector partnerships, and smart policies can protect Europe’s democratic processes in a fast-changing digital world. Co-organised by Horizon Europe projects AI4TRUST AI-CODE, AI4Debunk, TITAN, veraAI, and ELLIOT, with the support of EU DisinfoLab and EDMO. 
  • 17 December: Media & Learning Wednesday Webinar about Lines of speech: hate, harm and the laws across borders.
  • 23-24 January: The Political Tech Summit, held in Berlin, offers an opportunity for political professionals working at the intersection of tech, campaigning, and democracy to exchange knowledge and discuss fresh perspectives shaping digital politics. 
  • 24 January: CDPD 2026 Call for Papers. 
  • 23 January-June 2026: The Cyber for Good Media Academy will take place with the mission to protect and better equip journalists against interference and manipulation in the digital space, with a focus on OSINT and cybersecurity. Applications for this event open on 3 November and close on 5 December, with the program beginning on 23 January. 
  • 16-17 February: The DSA and Platform Regulation Conference will take place at the Amsterdam Law School, to reflect on the DSA and European platform regulation, providing an opportunity to discuss its broader legal and political context, through the overall theme of platform governance and democracy.     
  • 8-10 April: The Cambridge disinformation summit is expected to gather the world’s leading scholars, professionals, and policy-makers to explore interventions on systemic risks from disinformation.

Jobs 

Did you find a job thanks to the listing in this newsletter? We’d love to know – please drop us a message!

SUBSCRIBE TO THIS NEWSLETTER !

Have something to share – an event, job opening, publication? Send your suggestions via the “get in touch” form below, and we’ll consider them for the next edition of Disinfo Update.