Dear readers,

Summer is well underway, and before we take a short break, we leave you with some summer reading. True to form, this edition stays focused on enforcement.

One highlight is a recent ruling by the Court of Justice of the European Union, which reaffirms that the application of sanctions against Russian propaganda admits no exceptions. The Court confirmed that making content available via streaming through a sanctioned entity constitutes a sanctions violation and may trigger criminal proceedings.

This is one of many topics we will be exploring at our annual conference, which will mark a major milestone for the semester ahead. It’s also a reminder that Europe’s legal arsenal, when actually used, is already powerful enough to limit the spread of propaganda that threatens our democracies. To get a taste of what else is in store, have a look at our recent #Disinfo2026 programme highlights, and make sure to grab your tickets!

The DisinfoLab team is now taking a well-deserved summer break. Our newsletter and webinars will be on pause until early September, when we return refreshed and ready for the busy months ahead.

The entire DisinfoLab team wishes you great reading and a wonderful summer.

FOREIGN INTERFERENCE & HYBRID THREATS

Sanctions enforcement against Russian propaganda has been clarified by the EU’s top court, which refines understanding of enforcement of the EU’s criminal law liability for breaches of EU sanctions. The EU Court of Justice ruled that the bloc’s RT ban applies to free, donation-funded websites as well as traditional broadcasters, arguing that a commercial motivation for websites publishing sanctioned content was not necessary to be found in breach of the law. In the meantime, RT resurfaced on X under a new handle to bypass the blockade and rack up millions of views. Simultaneously, a new investigation exposed “Roska Bridge”, a sophisticated Russian operation exploiting structural vulnerabilities in decentralised architectures. The campaign weaponises technical gateways like Brid.gy to cross-post propaganda waves automatically across Mastodon and Bluesky, systematically evading conventional content moderation. This operates alongside “Hahaganda”, a separate Russian psychological campaign using weaponised mockery across Europe, and a Chinese-linked covert network deploying hundreds of automated dating accounts to seed division ahead of Taiwan’s elections.

  • The EU Court of Justice ruled that the bloc’s RT (formerly Russia Today) ban applies to free websites (Eunews)
  • RT evades sanctions: new accounts bypass EU blockade (NewsGuard)
  • Roska Bridge: How a pro-Russian IMS exploits vulnerabilities of decentralised platforms to spread propaganda (CheckFirst)
  • Hahaganda: how Russia seeks to reinforce disinformation narratives in Europe through mockery and parody (Maldita)
  • Chinese network launches hundreds of fake dating accounts to influence the next Taiwanese election (NewsGuard)

NATIONAL AND INTERNATIONAL RESPONSES

At national level, discussions emerge to strengthen current regulations and top courts are mobilised for enforcing them. In France, QuotaClimat appealed to the Conseil d’État after media regulator Arcom failed to penalise CNews over unchallenged broadcasts claiming global warming stems from astrophysical “pluriverses”. Concurrently, French Prime Minister Sébastien Lecornu proposed a new legislative package to triple criminal penalties for election disinformation and expand fast-track judicial takedown orders (référé électoral) to all ballots. Meanwhile, a widening UK government boycott of X saw the Attorney General and the Department for Culture, Media and Sport (DCMS) remove their official accounts over incitement concerns following violent June unrest in Belfast and Southampton. Simultaneously, the DCMS launched a Green Paper consultation to legally mandate platform prominence for trusted news. Globally, Canada introduced The Safe Social Media Act (Bill C-34), creating a Digital Safety Commission to enforce an under-16 social media account ban and legally compel platforms and AI chatbots to label and mitigate systemic online harms, including deepfakes and hate propaganda.

  • QuotaClimat takes case to Council of State after regulator refuses to act on climate disinformation (Le Monde)
  • National strategy and legislative proposals to combat foreign electoral interference (Services du Premier Ministre, France)
  • UK Attorney General tells department to stop using X amid disinformation concerns (The Guardian)
  • UK Government “green paper and consultation”: A new strategic direction for UK media (UK Parliament, Hansard)
  • Canada moves to regulate online safety: Understanding the Safe Social Media Act -Bill C-34 (Borden Ladner Gervais)

PLATFORM GOVERNANCE & RESEARCH

Research by regulators and academics once again shows how platforms are failing to mitigate their systemic risks. An Austrian Digital Services Coordinator (KommAustria) study uncovered over 634,000 fraudulent ads on Meta’s network, generating more than one billion impressions by impersonating trusted media brands. Similarly, another study found malicious actors weaponising Telegram channels of Spanish public entities and media outlets to run unauthorised financial and betting scams. Meanwhile, studies on X’s Community Notes indicate the crowdsourced model struggles with polarising content and leaves election disinformation undermoderated, a deficit reinforced by a Harvard study of 29 fact-checkers warning against treating AI “Note Writers” as reliable human moderation replacements. Ultimately, a separate analysis warns that countering authoritarian disinformation requires macro strategies reaching far beyond basic platform governance.

  • Analysis of the fraud ecosystem in online advertising on Meta platforms (ÖIAT / KommAustria)
  • Telegram promotes financial scams and unauthorized betting operators in the channels of Spanish media outlets and public institutions (Maldita)
  • Community Notes undermoderate polarising content by design (Science Advances)
  • Fact-checking at a crossroads: fact-checkers’ perspectives on Community Notes, AI integration, and design recommendations (Harvard Kennedy School Misinformation Review)
  • Countering authoritarian disinformation requires more than platform governance (Tech Policy)

EU POLICY, REGULATORY DEVELOPMENTS & SOVEREIGNTY 

The European Commission is warning again that Meta is failing to mitigate systemic risks under the DSA. The European Commission issued formal preliminary findings ruling that Facebook and Instagram’s addictive elements, including infinite scroll, autoplay, and push notifications, breach safety laws by harming minors’ mental well-being. Concurrently, the Commission’s second annual DSA report flagged persistent threats from disinformation, election integrity, and AI content across large online platforms and search engines (read more on the Brussels Corner). Seeking to reduce external dependencies, the Commission migrated its communications to W Social, an AT Protocol-based European network, sparking debate among analysts over whether relying on closed-source ecosystems undermines digital sovereignty. Policy friction is also rising over the AVMSD revision; EU DisinfoLab warns that a proposed “media exemption” clause could shield rogue actors from moderation, opening a dangerous loophole for systemic disinformation. Meanwhile, the European Parliament demanded baseline digital literacy to help citizens spot manipulative architectures. Regulatory scrutiny has concurrently expanded to e-commerce; Germany’s Digital Services Coordinator launched formal DSA compliance proceedings against eBay for not making required seller details available and failing to provide a user-friendly process for reporting illegal content. Additionally, Meta faced a formal European Parliament challenge by MEP Alexandra Geese over algorithmic over-moderation that allegedly “shadowbans” civil society activism protecting women’s health and LGBTQIA+ rights.

  • Commission preliminarily finds the addictive design of Instagram and Facebook in breach of the Digital Services Act (European Commission)
  • Second report on systemic risks on Very Large Online Platforms and Search Engines under the DSA (European Commission)
  • Europe’s W Social Bet Tests its Vision of Digital Sovereignty (TechPolicy)
  • European Parliament calls for digital education so people can recognise disinformation, AI content and addictive design (2EU Brussels)
  • ⚠️Policy Alert: A revived “media exemption” in the Audiovisual Media Services Directive (AVMSD) could open a fresh loophole for disinformation actors. Read here our concerns and suggestions on the “media exemption” in the AVMSD.
  • German Digital Services Coordinator calls on eBay to uphold user rights under the DSA (Value Added Resource) 
  • EU monitors Meta’s online censorship of European activists following parliamentary challenge (EUNews)

AI DISINFO WATCH 

AI Summaries: still a black box and a disinformation risk. AI-generated summaries remain one of the biggest challenges to information integrity, particularly for researchers. Recent findings point to a persistent lack of transparency in how these tools select sources, and a troubling tendency toward “fact laundering”, retrieving accurate information but attributing it to secondary sources rather than the primary ones. Fact-checkers are acutely aware of this problem: at GlobalFact 2026, discussions repeatedly returned to AI’s dual role, both as a useful tool in verification workflows and as a source of bias and hallucination (and Full Fact offered one example of the former). One promising fix comes from NewsGuard, which has built NewsGuard AI, a chatbot that draws exclusively from sources it rates as reliable, with fully attributed responses. Users themselves seem increasingly aware of these limitations too: according to the latest Pew Research Center survey, 60% of Americans now read AI-generated summaries, but only 29% actually trust the information these tools provide.

  • Fact laundering: AI research tools find facts and lose the source (Velora)
  • When AI systems select the news (Agora Digitale Transformation)
  • Fact-checkers wrestle with how to minimize AI’s problems (Poynter)
  • Full Fact is battling AI-generated election content with AI tools of its own (Nieman Lab)
  • Why we just launched “NewsGuard AI” (NewsGuard)
  • Americans and AI 2026: Chatbots, smart devices and views on impact (Pew Research Center)

Explore more in our recently updated AI Disinfo Hub, where we gather the latest research, tools, and analysis on AI-driven disinformation.

CRISIS AND CONFLICTS 

When a crisis strikes, bad actors rapidly exploit public fear. This is clear from conspiracy networks blaming HAARP program for the twin earthquakes in Venezuela and European heatwaves to stoke anti-US sentiment, as well as political figures stripping an oncology study of context to falsely claim sunscreen causes skin cancer, undermining public health. On a larger scale, at the UN General Assembly, the Secretary-General’s benchmark report and a subsequent EU statement, both centered on preventing conflict, genocide, and atrocity crimes, identified disinformation and hate speech as key drivers of conflict and atrocity risk, alerting that early warning signs are too often ignored. This cognitive warfare is underscored by research exposing Russia’s systemic Foreign Information Manipulation and Interference (FIMI) ecosystem, which leverages generative AI and localised laundering models to derail Ukraine’s integration into Europe.

  • Blaming the deep state for earthquakes (NewsGuard)
  • What? I should skip the sunscreen? (NewsGuard)
  • Amid rise in conflicts, impunity for International law violations, responsibility to protect more vital than ever, Secretary-General tells General Assembly (United Nations)
  • EU Statement – UN General Assembly: Responsibility to protect and the prevention of genocide (European External Action Service)
  • Beyond the battlefield: Russia’s information war against Ukraine’s European future (European External Action Service)

Explore more in our recently updated Conflict & Crisis hub, where we gather the latest research, tools, and analysis on AI-driven disinformation.

Brussels corner 

Soft law requires hard work. From the outset, it was clear that the EU Code of Conduct on Disinformation risked wasting political capital and civil society resources that are desperately needed for the effective enforcement of the DSA. The areas covered by the Code are important, but can a voluntary instrument credibly address problems rooted in platforms’ business models? In our latest blog post, we look at the Council of Europe’s Guidance Note on best practices for self- and co-regulatory approaches to content moderation – a document approved by all EU Member States, but largely ignored during the DSA negotiations. 

All questions and no answers. At the European Parliament LIBE Committee meeting on 24 June, MEP Birgit Sippel (S&D) asked Digital Commissioner Henna Virkkunen a question central to DSA enforcement: have the DSA’s sanctions and fines changed companies’ behaviour, or are they still spreading illegal content and disinformation contrary to EU legislation? Unfortunately, the Commissioner either forgot or chose deliberately not to answer.

Systemic risks’ list. Article 35(2) of the DSA requires the Digital Services Board and the European Commission to prepare an annual report on the most frequent systemic risks caused by the design and functioning of very large online platforms and search engines, and to identify best practices for mitigating those risks. The second report was published on 2 July 2026. Interestingly, it readily admits to not, yet, identifying “best” practices. Instead, it promises that “in light of accumulating experience with DSA implementation and enforcement in practice, future editions of this report will also aim to identify evolving best practices for the mitigation of systemic risks.” 

YouTube? Let me tell you about some other VLOPs. In response to a parliamentary question from MEP Dan-Ştefan Motreanu (EPP), Commissioner Virkkunen said the European Commission is monitoring YouTube’s compliance with the DSA and can use its investigative powers to address any non-compliance. After a few lines on YouTube, however, the response shifts to enforcement examples involving Temu, Shein, AliExpress and TikTok. Oddly, none of the examples cited by the Commission are American.

Intelligence staff question unaddressed. MEPs Sandro Ruotolo and Nicola Zingaretti (S&D Group) asked the European Commission, in a parliamentary question, whether it exercised oversight over or had any information concerning Meta’s “election operations centres”, active during the 2018 and 2022 Italian elections and the 2024 European elections, and whether it intended to assess the role of former US intelligence agents, who partially staffed the centres, in shaping and applying Meta’s political content policies.Commissioner Virkkunen replied that the Commission is aware of the centres based solely on information provided by Meta and referred to the Commission’s non-binding Election Guidelines, leaving the question of the potential influence of staff with US intelligence backgrounds on Meta’s political content policies unanswered.

A Commissioner who doesn’t answer. One could be forgiven for wondering whether there is a pattern in the answers given by Commissioner Virkkunen on questions about online regulation. MEPs from several political groups have raised issues ranging from platform practices and YouTube’s recommender algorithm to Meta’s election operations centres, the downgrading of LGBTQI+ accounts on Instagram and scam advertisements. In each case, the Commissioner referred to broader investigations, enforcement actions or voluntary guidelines without fully addressing the specific questions raised.

Explore the full Brussels Corner on our website.

🧡 One thing we loved

A major legal victory for disinformation researchers. After three years of intensive litigation, a U.S. federal court officially dismissed a high-profile, politically motivated lawsuit brought by America First Legal on behalf of the right-wing outlet The Gateway Pundit. Prominent researchers, including Renee DiResta and Kate Starbird, were falsely accused of running a “censorship industrial complex” for their work tracking election and public health misinformation.

While the targeted harassment campaign sadly caused real-world collateral damage, including the shutdown of the Stanford Internet Observatory, the court’s dismissal represents a significant vindication for the research community and a major setback for those behind the lawsuit.

🤝 Community meetups

Our informal community meetups are taking a short summer break.

If you’d like to join fellow researchers, journalists, policymakers, technologists and civil society practitioners at future events in Brussels and other European cities to exchange ideas, make new connections and enjoy an informal evening together, simply reply to this email with the city or country you’re based in, and we’ll keep you updated about meetups in your area. 

🧰 Tools & learning

The invisible hands of Polymarket. OSINT trainer and toolmaker Henk van Ess has built a free tool that monitors Polymarket’s biggest markets for signs of manipulation, identifying potentially manipulated betting activity to make a market look busier or more one-sided than it actually is. A timely resource, given growing scrutiny of prediction-market integrity and opaque influencer payments. Try it here

Ethics in the fight against disinformation. As part of the EU-funded PERiSCOPE project, EU DisinfoLab published a new self-paced training course developed  for journalists, educators, civil society organisations, AI practitioners and others working to protect information integrity. It explores the ethical challenges of detecting and responding to disinformation – including fact-checking, OSINT, AI tools and platform accountability – through six practical modules featuring case studies and reflection exercises. The course is free to access, with no prior technical expertise required, and a launch webinar recording is also available.

💬 EU DisinfoLab webinars

Our webinar series is also taking a summer break ☀️

Thanks to everyone who has joined us over the past months. We truly appreciate your participation, engaging discussions, and insightful questions.

We’ll be back in September with a new season of webinars. Until then, we wish you a wonderful and relaxing summer.

In the meantime, browse our webinar library and catch up on any sessions you may have missed. Happy watching!

🗓️ Events on our radar

🤝 Jobs & opportunities

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.