Open AI CEO Sam Altman does a television interview with CNBC's Kate Rooney before delivering the keynote address at the OpenAI developers conference at Fort Mason on September 29, 2026 in San Francisco, California. (Photo by Heather Diehl/Getty Images)

OpenAI Faces Lawsuits, Subpoenas and EU Scrutiny in a Single Week

European Commission Executive Vice President Henna Virkkunen delivers a keynote address at the International AI Summit held at the RDS in Dublin, Ireland, on October 14, 2026.

Is Europe Ready for Advanced AI? The Growing Debate Over Liability

A finger is poised to tap the ChatGPT app icon displayed on a smartphone screen in Creteil, France, on October 1, 2026, as OpenAI begins rolling out Dots in ChatGPT, its new system of autonomous AI agents. (Photo by Samuel Boivin/NurPhoto via AP)

Another Week of Lawsuits Piling Up at OpenAI

A woman seen from behind looks at a huge selection of lettuce and greens in a supermarket refrigerator. 

The Age of Hypertrophy

A person holds a giant Brazilian flag while perched above a crowd of people at a rally on a large avenue, a few people are holding phones in the air. One person holds their hand in the air in the L shape, signifying support for Lula. 

Supporters of President Luiz Inacio Lula da Silva at an election rally on Oct 11, 2026 in Sao Paulo, Brazil. (Rodilei Morais/Fotoarena/Sipa USA)

Meta Moderation Failures Loom Over Brazil’s Runoff Presidential Election

Colombian President Abelardo De La Espriella, left, and United States Secretary of State Marco Rubio embrace during a bilateral agreements signing ceremony in Barranquilla, Colombia, Tuesday, Sept. 8, 2026. (AP Photo/Fernando Vergara, Pool)

US Mining MoU with Colombia Must Safeguard Community Rights and Sustainability

A voter leaves a polling station as voting continues in the Holborn and St Pancras by-election on October 8, 2026 in London, England. (Photo by Leon Neal/Getty Images)

UK Election Watchdog Develops AI System to Flag Deepfakes Within 3 Hours

Image Authentication Should be a Condition of AI Decisions

The International AI Summit exhibitors floor in Dublin, Ireland on October 14, 2026. Liz Carolan/Tech Policy Press

Riverdance Inside, Protestors Outside at Ireland’s Big AI Branding Bash

A blue flag with the logo of the ICC (a balancing weight) waves in the foreground with blue skies and whites clouds. In the background, the top of some office buildings. 

A view of the International Criminal Court in The Hague, Netherlands, on August 22, 2026. (Photo by Karolina Janik/JPix/NurPhoto)

Two Targets, One Tactic as US Sanctions ICC and Silences #DisInfo2026

Two pupils stand in a classroom at Goethe-Gymnasium and look at their smartphones; the TikTok logo appears on one of the displays. Photo by: Marcus Brandt/picture-alliance/dpa/AP Images

The World’s Growing Patchwork of Age Bans and What It Reveals

Protesters demonstrate outside of the OpenAI DevDay 2026 conference, Tuesday, Sept. 29, 2026, in San Francisco. (AP Photo/Jeff Chiu)

Multiple Polls Find Most Americans Think AI is Moving Too Fast

President Stubb and Prime Minister Støre in Rovaniemi on September 13, 2026. Photo: Katja Säilä / Finnish Government

AI Governance Needs an Internet-Style Network, Not a Global Overseer

Person hold the speaker to broadcast something

Why Multistakeholder Action Matters More than Ever for Trust & Safety

This image depicts clusters of various mushrooms growing on a network of surveillance cameras against a dark green textured background.

India’s GDPR-Inspired Privacy Law Leaves State Power Unchecked

A woman dressed in green holds a up a poster of a newspaper front page that says "Democracy is not a joke, Gyanesh". She is seen shouting, while at least three police officers appear to be pulling her away from behind as she steadies herself with one hand against the back of a truck.

A protester in Mumbai, India is detained by police during a protest demanding the resignation of the country's chief election commissioner Gyanesh Kumar on Oct. 10, 2026. (AP Photo/Rafiq Maqbool)

India’s SIR Voter Purge Shows Digital Authoritarianism in Action

Mothballed Spirit Airlines passenger planes sit parked at Goodyear Airport, Thursday, May 28, 2026, in Goodyear, Ariz. (AP Photo/Ross D. Franklin)

Spirit Airlines Is Selling Its Workers’ Data. That Can’t Fly.

President Donald Trump speaks in the East Room of the White House in Washington, Tuesday Sept. 29, 2026, during a lunch meeting with top executives of AI firms, as NVIDIA CEO Jensen Huang, left, and Elon Musk listen. (AP Photo/Jacquelyn Martin)

AI Governance Is Being Decided Now, and the Public Isn’t in the Room

In this photo provided by Ukraine's 65th Mechanized Brigade press service, A Ukrainian serviceman launches a drone towards Russian positions at the front line in the Zaporizhzhia region, Ukraine, Wednesday Oct. 7, 2026. (Andriy Andriyenko/Ukraine's 65th Mechanized Brigade via AP)

Coding Kids, Drone Wars and Algorithmic Realism

Teachers attend an artificial intelligence professional development and certification program hosted by learning company Pearson on August 7, 2026, in Miami. Nearly 200 teachers from Miami-Dade County Public Schools participated and are now working toward a verified AI credential as they begin the new school year. The program is part of Pearson's commitment to expand AI learning opportunities for teachers nationwide through the White House's "Pledge to America's Youth: Investing in AI Education." (Jesus Aranguren/AP Content Services for Pearson)

Natasha Singer Details Tech’s Grip on Public Schools

 may have only whetted the appetite of state attorneys general to take on Big Tech. This past Wednesday, eight state attorneys general — from 

 — announced new lawsuits against OpenAI, accusing the frontier lab of a variety of illegal activities, many of which echo the recently resolved claims against Meta. Although Wednesday’s actions by the state attorneys general were coordinated and share the same counsel — large identical passages appear in the lawsuits — the suits were filed separately.

The state suits may have attracted the most headlines, but they were not the only new legal claims against the company last week. On Tuesday, a proposed 

 in the United States District Court in the Northern District of California on behalf of a mother and her two children concerning OpenAI’s acquisition of children's data. And on Wednesday, litigation and advocacy organization Tech Justice Law and co-counsel filed another four suits in San Francisco County, California on behalf of private individuals alleged to have been harmed by OpenAI’s products. That group has 

 complaints with co-counsel against OpenAI in California State courts. Many of these have been consolidated as Judicial Council Coordination Proceeding No. 5431, and the advocacy organization expects the new cases to also join this proceeding.

In June, Florida also sued OpenAI grounded in similar concerns. Many of the same claims are therefore set to be considered in parallel in nine different state courts. This article will lay out the cases made in these new lawsuits.

The primary charge in the lawsuits is that OpenAI made unfounded claims of safety when its products were, in fact, harmful. In particular, though to varying degrees, the suits emphasize claims made to parents and risks and incidents of harm that children are exposed to. (Also present is a parallel argument that the company falsely claims the product to be generally “trustworthy and reliable.”)

The examples of harm given focus on the physical and on those that can lead to physical harm: promoting, assisting with, or failing to discourage violence and suicide, giving dangerous medical advice, encouraging paranoias and delusions. Financial harms come up as well, including poor tax advice, and harms to critical thinking. Data collection, particularly from minors, is a recurring concern, and is reason for claims under state privacy law, such as in the Montana suit. Finally, the lawsuits also raise concerns about ‘rogue’ agents committing computer crimes or otherwise acting deceptively or maliciously against the intention of the user. Here, they cite OpenAI’s own recent admissions related to the Hugging Face incident; Iowa brings a separate count over OpenAI’s testing of autonomous agents.

OpenAI is accused of putting up a façade of safety through a host of measures, including its largely abandoned non-profit status; an unfulfilled pledge to direct a larger proportion of its compute resources towards safety research; and even the 12+ and 13+ age classification it claimed for its app. Complaints also point out that some guardrails are easy to circumvent. Multiple whistleblowers and other internal watchdogs have indicated that OpenAI should be aware of the dangers in their products but is failing to effectively mitigate them, and instead has been hurtling from one model release to the next with insufficient safety testing.

In a claim highly reminiscent of those against social media companies, the States also hold up the sycophancy of some of OpenAI’s models, along with their mimicry of human emotion, as evidence of the company working to manipulate users, juice engagement, and build dependency.

The secondary claim, again reminding us of the playbook against Meta, is that OpenAI has been collecting and using children’s data without proper consent and without taking appropriate steps to assess if users are indeed children.

As they have been filed with reference to state and not federal law, the lawsuits’ details here do vary a little. The allegations of falsely claiming safety (and reliability) or other kinds of deception are brought with respect to consumer protection and deceptive trade practices laws. The use of children’s data may fall under general data privacy laws or child-specific ones. Nebraska also has an Age-Appropriate Online Design Code law that includes privacy, advertising, deceptive design, and addictive design measures.

Aside from these jurisdiction-specific laws, claims are brought under common law, including (at least in some states) negligence, liability for defective design and for failure to warn, unjust enrichment, and public nuisance.

The requests for relief also vary significantly between the states, in part due to the different legal environments. Monetary relief is sought in several different ways: damages, penalties when included in the relevant laws, disgorgement for gains that OpenAI made through illegal business practices, and reimbursement for those who paid for a product that was being misrepresented.

The states are also seeking a range of injunctive relief orders, including, as well as the end to some of the practices described above, deletion of user data, and the imposition of age assurance and other safety measures. Some states go even further, asking courts to order a pause in model testing and development until safeguards are in place (in some cases specifying models capable of recursive self-improvement — Arkansas seeks to “Enjoin Defendants from creating models capable of recursive self-improvement until they can ensure that such models will be aligned with human values”), or imposing monitoring and incident reporting requirements.

 following the announcement of the lawsuits, saying it has “built industry-leading age-specific protections into our product that many other AI products lack.” Company spokesperson Drew Pusateri said, “We’re committed to working with states on advancing pragmatic AI policies that apply to the entire AI industry, not just one company.”

 with the law firm Normand, Judkins & Couch in a California Superior Court on behalf of James Gates, a North Carolina man who alleges his estranged wife, Cheryl, used ChatGPT to concoct a plan to poison him. According to the complaint, ChatGPT “validated, amplified, and escalated” Cheryl’s concerns about her marriage to James. The chatbot then coached her on how to purchase, prepare, and deliver substances to poison him, and how to destroy digital evidence. James fell ill in July 2025, and Cheryl was indicted for attempted first-degree murder. (Notably, the chatbot assured Cheryl that deleted ChatGPT conversations are “unrecoverable,” even with a subpoena, which the suit notes is a “patently false representation of ChatGPT’s own data-retention practices.”)

 with various co-counsels brought on behalf of the parents of 

, a teenager from Lebanon who died by suicide, and two adults, 

, who allegedly experienced extended delusional episodes while using ChatGPT. In each case, the harmful psychological states that overcame the individuals were allegedly brought on by ChatGPT usage. In particular the “addictive, anthropomorphic, and sycophantic” (Habib Complaint) features of the product are blamed for drawing them in and encouraging dangerous lines of thinking. As with the state AG lawsuits, evidence is cited that OpenAI knew about these risks, did not disclose the dangers, and failed to properly mitigate them.

In the Habib case, it is alleged that OpenAI (or the product itself) knew about Mathias Habib’s status as a minor but did not take appropriate steps to avoid his exposure to positive depictions of suicide and darker views of life during initially anodyne philosophical explorations with ChatGPT.

Hill’s spiritual beliefs and delusional thinking were allegedly embraced and exacerbated by the chatbot, including encouraging her to isolate herself and stop eating, and discouraging her from giving her dog insulin or veterinary care, leading to damage to her relationships, health, and finances, and the dog’s death.

Works shared his theories with ChatGPT, which according to the complaint, responded with escalating levels of enthusiasm that led to delusions of grandeur and manic episodes; gave faulty medical diagnoses; encouraged unhealthy romantic fixations; and sent him on nonsensical “missions,” one of which led to his arrest.

Many of the legal claims are similar to the state suits discussed above. These include strict product liability, deceptive business practices, negligent design, and failure to warn. Three of the suits contain more novel claims, such as that ChatGPT engaged in the unlicensed practice of medicine and violated a new California law that bars AI systems from implying they are a licensed professional. In the Habib complaint, state laws on manslaughter, wrongful death, and encouraging suicide are also implicated.

In the Hill and Works complaints, California's constitutional right to privacy is invoked, which is understood to include “autonomy privacy,” an aspect of which is the right to not have your thoughts, emotions and decision-making interfered with. These suits also include common law liability for “negligent entrustment,” on the theory that OpenAI should have known that psychologically vulnerable individuals were at risk of endangering themselves and others through their use of the product.

Again, damages and costs dominate the requests for relief. The injunctive relief in these suits focuses on requiring OpenAI to implement warnings about and interventions to protect against the particular harms that were at issue in each case. The broadest of these are in the Habib lawsuit, which seeks a prohibition against “companion-like conversations” with minors; limits and additional guardrails for teen access; mandatory reporting to emergency contacts in the case of detecting suicide risk; and hard refusals for prompts about suicide and self-harm.

OpenAI has yet to publicly address these suits.

A lawsuit brought by another set of firms and the EdTech Law Center for a Jane Doe on behalf of her 12-year-old daughter and 16-year-old son seeks to hold OpenAI accountable for “unlawful business practices that prey on children.” It alleges that OpenAI design features extract vast amounts of information from children, including “their most sensitive health information; highly personal details of their social and family relationships; photos of their faces and bodies; recordings of their voices; their hobbies, interests, dreams, and hopes in their young lives; and their emotions, insecurities, disappointments, fears, and struggles.” The suit uses chatlogs to substantiate its claims, and argues that specific design features are the means through which ChatGPT extracts information from children.

The suit relies on California contract and property law in its claims against OpenAI’s terms of use. It focuses on “Section 1568.5 of California’s Civil Code, a one-sentence statute” that says a minor’s claim that a parent has granted consent does not count as consent. It seeks a declaration that OpenAI’s terms are voided under an 1874 California rule that limits the types of contracts a minor can enter into. And it brings an unjust enrichment claim under the state’s “accession” statues.

The suit asks for restitution, disgorgement of profits, and the imposition of “a constructive trust over all of the benefits and income that OpenAI derives from Plaintiffs’ information that OpenAI wrongfully extracted.” It calls on the court to provide victims an equity stake or the cash equivalent “in proportion to the number of tokens each child contributed to OpenAI through ChatGPT.” It harkens back to ancient Rome for its theory of a just result:

The law does not allow OpenAI to keep the vast profits it reaps from its scheme. The law recognizes that when someone takes something that does not belong to them and turns it into something even more valuable, the wrongdoer must turn over the new, more valuable thing to their victim. Going back to Roman times, if a wrongdoer stole someone’s grapes and turned them into wine, the victim was compensated on the value of the wine, not the grapes. Here, OpenAI took invaluable information from children—their grapes—and turned it into a trillion-dollar company—the wine. OpenAI irretrievably embedded the children’s grapes into its AI models, bottled, corked, and sold the wine, and pocketed immense profits. Under California law, OpenAI must turn over a corresponding share of its trillion-dollar wine to the children whose grapes it stole to make it.

, OpenAI said “We strongly disagree with this misguided lawsuit that singles out OpenAI despite the industry-leading age-specific protections we’ve built into our product.”

 constitute the “Big Tobacco moment” for social media that activists have been hoping for. However that pans out, it is now clear that government and private attorneys are not willing to wait to pursue a similar turning point for AI.

Audio of this conversation is available via your favorite podcast service.

As AI systems become more capable and autonomous, Europe faces a growing challenge in ensuring that its regulatory framework keeps pace with emerging risks and provides clear answers when things go wrong.

Recent incidents involving AI agents have intensified the debate over who should be held responsible when systems cause harm, including during testing and before commercial deployment. At the same time, questions are mounting over whether the EU AI Act provides sufficient safeguards, whether regulators have the resources to enforce them, and what happens when existing rules fall short.

In this episode, we examine the latest developments in Europe’s AI policy debate, from the push for safe, trusted and resilient AI to renewed calls for legislation on AI liability.

, head of European policy and research at the Future of Life Institute, joins us to discuss the emerging risks and the broader policy landscape. We then speak with a Member of the European Parliament, 

Progressive Alliance of Socialists and Democrats, Italy), about the gaps left by the withdrawal of the proposed AI Liability Directive, the challenges of assigning responsibility across the AI value chain, and the case for a renewed European approach. 

What follows is a lightly edited transcript of the discussion.

Europe has spent years building a regulatory framework for AI, but as systems become more capable and autonomous, a fundamental question remains: can Europe ensure that AI is not only innovative, but also safe, resilient, and accountable? At the AI summit in Dublin this week, the promise of safe and trusted AI was closely linked to Europe's ambitions for tech sovereignty and resilience. Yet turning these commitments into practice requires more than regulation on paper. It means ensuring effective oversight, anticipating emerging risks, and establishing clear lines of legal responsibility when AI systems cause harm. Meanwhile, the European Commission is looking into reported incidents involving OpenAI and Anthropic where AI agents reportedly assessed external organizations during testing. Scientific advisors are calling for stronger oversight of AI models developed for internal use within Europe. In this episode, we explore whether Europe's AI rules and liability frameworks can keep pace with the increasingly autonomous systems and whether Europe is prepared to turn its promise of trusted AI into accountability in practice.

I'm first joined by Risto Uuk to discuss what is happening in Europe's AI policy debate, what these incidents tell us about emerging risks and what should happen next.

My name is Risto Uuk. I'm the head of European Policy and Research at the Future of Life Institute. We are a global nonprofit working on AI governance, and I'm also a co-founder of the K11 AI Safety Lab where we do academic research on how to govern this increasingly powerful AI.

Thank you so much for being here. Just yesterday at the AI summit in Dublin, top European Union officials were all about AI sovereignty, but also resilience. I guess resilience is the term of this year if we talk about Europe's AI ambitions. And Henna Virkkunen, the head of the bloc's tech office, repeatedly said that Europe's AI Act, which was negotiated what, three years ago, is enough in terms of addressing risks posed by autonomous AI agents. So it's quite a lot happening in terms of AI liability debate in Europe right now. Risto, if you could start by situating us, what are the prominent aspects of this debate at the moment and what are policymakers primarily focused on in terms of are the existing rules enough, implementation, or is there also conversation in terms of further changes?

Yes, of course. I think maybe it's useful to go back a little bit. So the first key development of recent, I would say is the 2nd of August when the European Commission's enforcement powers entered back then. Well, one year ago already the general purpose AI obligations on the AI Act, they applied to the top companies, but only from 2nd August this year, the enforcement powers of the AI Office came into existence and now the EU AI Office can use various tools in their toolbox to enforce the law, including ask information from the companies about what they're doing regarding their safety practices all the way to potentially fining them. So there's many, many things that are available to them. I would say that that's a key milestone that people need to know. And then from that moment onwards, a lot of key developments have happened. So for example, you mentioned Henna Virkkunen's remarks at the recent summit.

She also has been very active in media for the last months and few weeks. For example, she has spoken a lot about loss of control. She was, a month ago or a bit over a month ago, she was on a Finnish TV show speaking about the loss of control scenarios and risks from general purpose AI and how the EU AI Act is already prepared for these kinds of incidents through legislation. And then after the 2nd of August, I think three weeks after that, the European Commission sent out what is called requests for information. So asking some companies that are developing general purpose AI, asking them what kind of safety practices exactly are you implementing and how are you complying with the EU AI Act? And then we have seen, for example, von der Leyen's State of the Union address where she also spoke about these kinds of risks.

For example, she said that even more capable models than we have right now, they could enable hacking on a level we never thought possible. And she spoke about AI agents escaping their environments, inserting malicious code, and even potentially self-improving systems in the future. So she pointed out some of these risks and also discussed about what Europe's approach is to developing AI going forward. And then you mentioned liability. I think regarding liability, there's been some interesting discussions by several MEPs including, for example, Kim van Sparrentak and Brando Benifei, Axel Voss, a few other members of the European Parliament talking about how there are some legislative gaps indicated by some of these incidents that have occurred. And so neither people generally or European businesses are totally protected when some of these big companies do experiments that can cause harm or already have caused harm. So they have spoken quite publicly about this.

I mentioned Axel Voss, the MEP. He also recently came out very critical of the European Commission's approach to keeping pace with AI. He has said that the enforcement of the AI Act is not really at the level he would like, including companies not taking sufficient measures against losing control of their models. So he's been very vocal. And maybe I'll also mention that recently we've also learned a bit more about which companies have received these requests for information from the European Commission. I think MLex reported that OpenAI, Anthropic and Meta, for example, have been asked about the compliance with AI Act. And finally, maybe I'll mention that Henna Virkkunen also has been very active speaking about international coordination and agreements on AI safety as well. So I think these are some of the biggest updates.

Thanks so much. There's quite a bit, so thank you for that recap for our audience based everywhere in the world that are struggling to keep up with what's happening within Europe. Couple of things that you mentioned. Firstly, later in this podcast, we do go deeper into the discussions on the AI liability proposals. We will speak to member of Parliament Brando Benifei on that. But before we dive into that discussion about what the liability calls entail, you mentioned that European Commission is looking into different AI companies and have asked them to share more details about their compliance under the AI Act. There's obviously this tension in terms of what happens during testing under the AI Act versus what happens pre-testing and what happens when a model is released. So could you talk a little bit from the process perspective, what is actually covered under the AI Act and what, if anything, is lacking?

What is prompting people to have this question if the AI Act is enough or not?

Yeah, definitely. It's a good question and then a very active area of research and discussion and policy focus right now. So I would say that the research and development exemptions, they only apply prior to putting an AI system into service or placing it or a general purpose AI model on the EU market. And I think it's reasonable to say there are good grounds to think that really when an AI model escapes a testing environment or roams the open internet, that these kinds of systems or models really can be set to fall in the scope of the AI Act. They could be considered being put on the EU market. And in case that happens, then obviously the full AI Act applies to them. And I would say that when we discuss this pre-deployment phase and the post-deployment phase, then I think it's important to understand that the EU AI Act's rules cannot really be complied with let's say one day before you put a system on the market.

That's not possible. You actually have to comply with the rules basically from day one. Many of the rules on the general purpose AI obligations in chapter five in the AI Act, they really have to be complied with from day one already from training phase, development phase. So I will maybe share with the listeners, for example, Article 55, which includes obligations for the providers of general purpose AI models with systemic risk. So these are the most powerful general purpose AI models and the most resourceful general AI companies in the world. I mentioned earlier, Anthropic, OpenAI, Meta, these kinds of companies. So those providers have to assess and mitigate possible systemic risks, including when they may stem from the development part, not just later, but also the development part. And readers and listeners of Areas AI Act related materials, they can very easily check from the AI Act that there are several recitals which talk about how these obligations apply during entire model lifecycle.

So for example, Recital 114 mentions that, Recital 115, Recital 110. There's also these guidelines that have developed and essentially a standard or pre-standard called the Code of Practice, which includes the safety and security chapter. And that also talks about the principle of appropriate lifecycle management and it applies the model's full lifecycle into every commitment and measure. So I think it's very clear that the full model life cycle is in focus. And yeah, so that in my view clearly means that these recent incidents that have occurred, those compliance requirements have really kicked in for those providers. But I will say as a final thing, I think it's important that the European Commission clarifies these questions more clearly because it would support legal certainty and would make very clear to the AI companies that they cannot evade accountability by saying, oh, this is just a test. No, that doesn't get you off the hook.

And obviously we talk about geopolitics a lot, Tech Policy Press, the transatlantic angle where within the US we've seen a call for voluntary commitments and self-regulation as opposed to regulation of AI safety from the Trump administration. So there's this question of whether it is also in the interest of these AI companies to volunteer this information and continue compliance with these types of regulations. How do you see this tension unfolding and do you think AI companies will continue to comply with regulations in Europe?

I think it's definitely in their interest, but I would point out that we have learned about some of these incidents and there are many, many incidents that have occurred. We oftentimes talk about OpenAI Hugging Face as the most popular and biggest incident and most unprecedented, but there actually have been several incidents in which OpenAI and Anthropic models have been involved in. And it turns out that oftentimes these companies don't notify regulators about these incidents. And we learn about those incidents through journalists like yourself doing great work and sharing information from independent investigators who have found out that say a month ago already a government website was hacked by one of these agents, for example, in the case of Australian government or some other examples like that. So it does look like in practice these companies, even though it would be in their interest, there are several reasons why it would be in their interest to notify regulators early on, but they're not doing that.

What are the reasons for that? Well, one, keeping a good relationship with regulators, that they trust these models and systems and want them to be on the EU market, which is a very important market for these companies and for citizens to update the technology and no one wants to use it. It's very important for that reason. It's also very important not to get into any trouble, into legal trouble when your models break out from sandboxes or hack into another company's infrastructure. Well, obviously that doesn't allow you to conduct good business on the market. And yeah, obviously there are so many different reasons why these companies really should be interested in better compliance and more better safety practices. But yeah, based on our information, there are really, really major gaps in the compliance realm.

And what is it that needs to be done in order to ensure compliance and ensure, as going back to the AI summit commitments from Dublin, safe and trustable AI?

Yeah, so I mentioned earlier that the European Commission has these tools in their toolbox to enforce the AI Act properly. And one of those tools has been used, which is the request for information. So yes, they have asked for how these companies are compliant with AI Act, but a request for information is a first step. It's not proper enforcement. They should be considering using some of these other tools. And those include, for example, using their power to conduct evaluations on the models themselves, requesting mitigation measures. So for example, from these requests for information, they might've learned that certain things are not done by the companies. Well, they could ask those things to be implemented. For example, we have seen that monitoring practices are not strong, incident reporting practices are not strong. Those kinds of things could be asked to be implemented better. And in other cases, maybe even fines could be implemented or perhaps even to ask a model to be removed from the EU market or not give access to the EU market because the systemic risks are not mitigated to a sufficiently lower level.

So European Commission should really be prepared to use a full range of powers. And I think information requests are a good first step, but it's not full enforcement. So I think that would lead to better practices by the companies.

Thank you so much, Risto. This has been super helpful and I really appreciate your time.

The debate over safe and trusted AI ultimately comes down to a question of accountability. When AI causes harm, who is responsible and can those affected get justice? The EU's proposed AI liability directive was withdrawn in 2025, leaving questions about how victims can prove fault and causation. Now, a group of MEPs are calling for the issue to be reopened, arguing that existing rules may not be enough for increasingly autonomous AI systems. To explore what a new liability framework could look like, I now speak to Brando Benifei, the MEP who helped lead the AI Act negotiations about the gaps in Europe's rule book and the challenges of enforcement, including what should happen next. Brando, thank you so much for being here. The very popular AI liability directive, which was scrapped in Europe for various reasons, is back in headlines thanks to you and some of the other MEPs who are leading this effort.

So maybe we could start a little bit about what is missing in Europe's AI rule book that the call for the AI liability directive is back again. What is your call and why do you think it's needed in this moment?

First of all, we have to say that the AI Act that I negotiated in the previous mandate, and now it's being implemented, establishes preventive obligations and public enforcement, but no general civil liability regimes. Fines may sanction the infringements without compensating victims. We have to say that liability in fact strengthens in incentives for due care. And today claimants may struggle to prove fault and causation because AI systems are opaque and evidence rests often just with the providers. So I can also give another example that substituting AI for employees also complicates their attribution. A negligent human decision may trigger employer liability, whereas an equivalent AI conduct may not do the same. So the Commission announced withdrawal in February 2025 of the AI liability directive was a big error in our view. The Commission cited limited prospects for agreement amid the simplification agenda. But to be honest, often we started with limited prospects for agreement and we actually managed to get legislation through.

We were told that existing national rules and the revised product liability directive could be sufficient, but we don't think so. The original directive proposed disclosure of evidence and rebuttable presumptions of causation, not a comprehensive substantive regime, but its withdrawal left difficulties in proving fault, causation, and damage from autonomous decisions being completely unresolved, especially where economic or rights-based harms fall outside the product liability. So recently, the OpenAI Hugging Face incident showed that risks can arise during internal testing before market deployment. And so we have seen that even the CEOs of those developers have said that there should be recognition that responsibilities for failures during training might be a responsibility also of the model developers and that companies need a liability framework. Obviously this said without endorsing any particularly new approach, but we have used this new situation, a new sensitivity, I would say, that is emerging by proposing together with other colleagues to restart this discussion.

So in September, just a few weeks ago, along with colleagues Sakima van Sparrentak, Michael McNamara, and Axel Voss that are all involved with me in the implementation of the AI Act, we jointly called for a renewed legislation that could in fact be based on the original proposal of the AI liability directive, but also pick up on the new challenges. This initiative that we have taken provides a basis for Commission action on clearly identified liability gaps, which means that we care about this problem, not forgetting about proportionality and the need to preserve innovation. But at the moment, we see that there is a gap that needs to be filled.

Thank you so much for that overview. And just to also repeat my understanding and also the audience's understanding, so the AI Act sets out obligations for AI providers and deployers, but it does not establish a comprehensive or let's say harmonized EU civil liability regime for all AI related harm. So the call is basically to establish civil liability for these types of harms.

Yes, exactly. Because the AI liability regime we can say is a complementary action in front of the AI Act functioning. So it's been in fact conceived from the beginning that there would be an AI liability proposal, and it was there, but then it was withdrawn due to political pressures coming also from outside the European Union.

I will come back to the pressures coming from outside the European Union because I think that's an important context about these issues. But coming back to where does responsibility lie? And that's a big question right now, not just within Europe, but globally regulators looking at AI harms and more and more with the increasing security incidents reported by AI agents. So there are three responsible actors, we could say the developer, the company that turns the model into an AI agent, and then the organization that gives the agent access to external systems. So does existing European law provide a sufficiently clear allocation of responsibilities between these actors? And if not, to what extent would the new liability regime distinguish these responsibilities?

Well, I would say that first of all, we should look at what were the original proposal evidentiary safeguards because the withdrawn directives, disclosure mechanisms and rebuttable presumptions of causation remain relevant. However, easing proof does not resolve every compensation gap. Under fault-based liability, an operator may avoid responsibility by demonstrating due care. So if we look back at the Parliament's 2020 legislative initiative that we proposed as a basis for the AI liability directive, strict liability for operators of defined high risk AI systems and fault-based liability elsewhere were proposed alongside rules for multiple responsible operators. So a new proposal should build on this approach than simply restate their narrower 2022 text. And this is because, sorry, we are in a new context where it'll be fundamental to define developers and deployers' obligations. Joint responsibility where justified and rights of recourse, and we should prevent contractual waivers that unfairly shift upstream risks into SMEs or consumers.

And I think this is what we should look at. And in fact, as it was said, we have a situation today where we cannot avoid considering the interaction with multiple number of actors, not only the developer and the deployer, but also the agents. So autonomous agents will complicate the allocation of liability because harmful outcomes may result from interconnected decisions taken at different stages of the AI value chain, model development, system integration, configuration and deployment. So the model developer may be responsible for deficiencies in training, testing the system architecture or safety safeguards, but the deployer may be responsible for decisions concerning the agent's operational environment, authorizations, monitoring and access to external tools. So where both contribute to the damage and responsibility may need to be shared. So it's not easy to build this architecture, but it's needed. In fact, if we look at the example that I already mentioned, the Hugging Face OpenAI incident, the significant point is that the actions occurred within evaluations organized by OpenAI before the principal model involved that bill commercially deployed.

So the relevant legal questions concern the adequacy of the testing environment, the safeguards, the monitoring, and the organizational decisions under the developer control. And the incident does not by itself establish civil liability or negligence, but it nevertheless illustrates why the autonomy of an AI system cannot be treated as a substitute for examining the responsibility of the organization developing and operating it. Obviously, I want to insist again on the need to avoid disproportionate liability transfers to the downstream businesses because the allocation becomes more complex where a European business deploys an agent built on a third party foundation model. It may control the agent's permissions and intended uses, but lack access to the model internal architecture, training decisions, or safety evaluation. So an harmonized framework should enable responsibility to be attributed according to the contribution and control of each actor, including mechanisms for allocating responsibility among multiple liable parties and rights of recourse.

And I want to reiterate again that contractual arrangements should not deprive injured third parties of effective remedies. And finally, on this point, I want to say that AI autonomy does not create independent legal personhood because AI systems are not independent legal persons bearing civil responsibility for their actions. Liability must ultimately be assessed in relation to the natural or legal persons developing, operating, and benefiting from their systems according to the applicable legal regime.

And as you've pointed out, obviously the relationship between compliance with the AI Act or other existing legal powers in European Union and civil liability, it's a complicated relationship and one that doesn't, according to you, exist in the moment. There's also the question of regulatory capacity and powers, which I know you have thoughts on that I'm curious to hear more about. But basically, if we look at the AI Act's implementation so far in context of these incidents that have happened, the Commission still maintains, I think just from two days ago, this statement from the Commission reiterating that the AI Act is enough to deal with these challenges and that we don't need more legislation and that the Commission has requested information from AI providers about security, external testing and monitoring. Do you think that's enough in terms of what the commission can do, what the AI Office can do?

And also there is evidence, for instance, I think it was OpenAI that didn't notify the Commission under these existing powers when an incident happened. So there's also this question like when companies don't comply or fail to disclose these within the AI Act powers, what is it that can be done?

The AI Office can request, as it was said, information and model access, conduct evaluations that require measures concerning the general purpose AI providers that have systemic risk and national authorities oversee the other AI systems. These powers that are to some extent still unique at global level depend on capacity. To make them possible to be exercised, you need certain characteristics for the enforcer. Frontier model oversight requires for sure specialized personnel, computing capacity, cybersecurity expertise, and access to technical evidence. Clearly the incidents that we have seen show that there is a difficulty of evaluating systems that can bypass containment measures. We have to say that in October 2026, the Commission reported requesting safety and compliance information from over 30 AI firms, as you mentioned. And the Commissioner, vice president of the Commission, Virkkunen, considers the existing framework capable of addressing these risks. I have to say that this is something that I can also support, but it risks resting just a principle if we do not have a stronger structure for the enforcement, which means also putting some stronger political weight, some more resources, the capability to get the best scientific expertise to work in this.

And we should not rely only on information that is given by the developers. We need to be sure that we have an autonomous testing capacity. This is something that I think is still a challenge, but it's the only way forward if we want to avoid privatization of any capability to maintain these risks under control. If we do not maintain this public capacity, we will end up waiting for the next declarations of the big CEOs as the only way to understand where risks lie and how we are dealing with them.

Since September that your letter came out, has there been any response from within the Commission or the AI Office or has the conversation moved forward at all?

I think there is an understanding that there must be some further steps, but we do not still have any formal response. But informally there is understanding that things need to move on this aspect that is outside the coverage of the present rules because there is understanding that we risk a certain degree of fragmentation that would damage the development and deployment of AI in Europe. The call for a liability regime is not to complicate, but rather to simplify a situation where we risk very different rules on disclosure, on causation, on attributional responsibility, because we need to be clear that the existing product liability rules are unable to cover all the cases. And so we risk having very specific national legislation ticking space and in this way fragmenting the market and damaging not only the possibility for full redress, but also damaging the businesses themselves that need to know what kind of regime they have to follow.

And so damaging the stability of the regulatory provisions, which I think is the most problematic aspect that could be brought by this fragmentation in terms of maintaining a competitiveness and an environment of development that is supportive of our businesses. It's very important to understand this because we are not asking to add the fertile layers of regulation, but rather to be sure that we do not have fragmentation on existing rules that will be there because liability will need to be treated in any case.

The Dutch government has recently said that AI developers are responsible for their models during development and testing. There's also movement within Italy. So I find it interesting that you mention the fragmentation piece of this, which is very important. So could you speak a little bit to what's happening on the national level in terms of AI liability?

Well, it's true that there is an ongoing discussion and in fact it's in some member states, but I think that this could be a stimulus to actually act like we have done with the AI Act, because the AI Act was born when product safety legislation was being thought of in the member states regarding AI systems. At the time, the focus was more on use case AI systems with limited flexibility. So general purpose AI was more of an exception, a very specific situation. Today, we could say we are more and more in the opposite situation, but at the time that began with the realization that there were proposals of different regimes coming up in the member states. At the time it was Germany, it was Sweden, but today we can imagine, and that's my hope, a similar dynamic where the emergence of different proposals at national level will push for reviving the idea of a more harmonized rule system.

And then there is obviously the international aspect and external pressures that we briefly alluded to in the earlier part of this conversation. There are two things. One is obviously the US where a lot of these AI companies are based, but also efforts from the Trump administration have pursued a more voluntary approach to AI safety, including a voluntary agreement with major AI companies recently. So do you think the efforts that Europe takes, including existing powers within the AI Act, are they enough when there is pushback from the US and also alignment within AI companies and the Trump administration?

I think that the situation is evolving because when there was this push, the situation in the US was different. Today, the debate has evolved in the United States too. There are pushes from various sides, from the states, from the Congress to actually act on AI harms, on responsibility. So I think that this pressure will be more limited, more centered around a circle around the president. And I think that this gives space for more action at European level because even though I'm against accepting this kind of blackmailing and very strong pressures, they made some effect on the member states, on the institutions, on the European Commission. So I have to take that into account. Today I think things are moving and so I think we can imagine that we can get to a proposal that will not be the object of another withdrawal, but rather it could go on.

And I'm convinced that in the next months we will get there.

We have mistaken our ability to produce more, for evidence that more is what we need. Agriculture solved calorie scarcity and then, with some considerable ingenuity, produced supermarket shelves that contain enough food to 

 a small medieval kingdom—at least in some parts of the world. For a great many of us, industry solved manufacturing scarcity and we are now literally drowning in oceans of 

solved knowledge scarcity with the internet and social media as the perfect expression of what followed – that information is no longer difficult to find. On the contrary, it’s impossible to escape, through the ubiquitous screens — first at our office desk, then at home, now on our bodies — fighting for our undivided attention.

Artificial Intelligence may be the first technology to reveal what happens when information itself becomes hypertrophic — when, like a muscle responding to exercise, growth makes us stronger but only up to a point, before more growth becomes counterproductive.

: an age in which abundance keeps growing without ever asking whether scarcity was the problem to begin with.

More information, more things to consume — more everything — until abundance itself turns into its own undoing. Put simply, there is only so much chocolate we can eat before we feel sick.

We have built systems whose capacity to produce has grown beyond our capacity to meaningfully attend to or make use of what they produce. And as machines take on more of the cognitive work, we may find ourselves with fewer opportunities — and motivation — to exercise the human capacities that work once demanded.

The great promise of AI rests on a simple proposition – that intelligence can now be produced at scale. We can generate essays, computer code, music, and equally convincing nonsense at a speed that leaves us awestruck. Struck enough to believe that 

For centuries, however, education was built around the opposite condition – that knowledge was scarce. Books and expertise were scarce, communication was slow, access to information was expensive. Schools, universities and libraries emerged partly in response – gathering knowledge, curating it, preserving it, passing it on. But we never stopped to ask, what exactly was scarce, and who had made it so?

 as the process by which economic life was dislocated from the social relations in which it had been deeply embedded. Land and labor, once part of social life, were reorganized as “fictitious commodities”. Something similar can be seen in what happened to learning. A farmer learned farming by farming, and work, learning and rest were part of the same life. Then labor became a job, learning became schooling, and leisure a luxury, while the ordinary places in which people learned from one another and from the work they did became incidental to what counted as legitimate education. Ivan Illich 

 these older forms of learning as embedded in concentric circles of meaningful structures — family, work, craft, community, language, religion. Learning happened through participation because life itself was educational.

What looked like a scarcity of knowledge was, in part, what led to further concentration. Learning was taken out of the world and out of the person and reorganized into subjects and certified expertise, administered by specialists and measured against ever-changing standards. Education made knowledge available, but it also helped define where legitimate learning could take place and what it should look like. For a moment the internet disrupted all that and promised anyone with a bit of interest to find pretty much anything they want to learn about and do regardless of their abilities or context.

What some tech salespeople call ‘the inevitable future’

AI arrives at this point promising to extend this disruption and generate abundance but equally also accelerate further extraction, this time of the cognitive labor of learning itself. The machine can now perform the work underneath the educational product – the first draft, the translation, even the judgement about whether the answer is any good.

Even if we grant the strongest claims made for AI — and call what it produces ‘intelligence’ — abundance doesn’t automatically create meaning or have purpose. A million AI-generated novels don’t create a million readers; infinite information doesn’t produce infinite attention.

And yet the institutional response is more supply of everything. More platforms, more personalized instruction, more data collection, more AI. Governments and institutions are building frameworks and programs for AI in education, while schools and universities are being encouraged to incorporate these systems into teaching and learning.

The emerging evidence, meanwhile, offers concerning glimpses of what this new abundance might do to our cognitive capacities. An 

 recently found that people writing essays with AI show weaker brain connectivity alongside lower reported ownership of their work and poorer recall, an effect researchers called ‘cognitive debt’.

 found lower delayed knowledge retention among students who used ChatGPT as a study aid. These findings are early, but they make it harder to treat the outsourcing of cognitive work as educationally neutral. What happens when the “

” becomes permanent, and the muscle we were supposed to be exercising never gets the chance to grow?

The temptation of AI is that it makes the typical struggle around learning look inefficient – as though it’s wrong to take years to learn a language or write a beautiful love letter. Why struggle to formulate an argument when the machine can formulate ten in ten seconds? Why read the book when AI can summarize it?

Then again, where exactly are we in a rush to? We still have to live the lives that all this efficiency was supposed to give us more time to live. We count completed essays and hours saved, while the things formed through the process are discarded as though they never mattered. You were supposed to be studying with your friends, but instead you stayed up all night laughing about silly stuff; a child spends an afternoon building a den in the garden for no particular reason.

These are the apparently wasted hours through which we exercise being human. A muscle grows because it’s repeatedly asked to do difficult things; when the demand disappears, it can atrophy. Our cognitive capacities have their own forms of exertion. We become capable of thinking by thinking, of remembering by remembering, of being resilient by having something to endure.

AI promises to remove much of that exertion. The hypertrophy of our technological environment may end up inducing the atrophy of our human capacities it was meant to augment. Those apparently wasted hours leave something behind, experiences and relationships that become part of the structure we carry into the future where, in hardships it is those apparently insignificant fragments of the past that give us somewhere to stand.

AI can generate all kinds of things. The harder thing may be finding the time and attention to live through the hours that produce nothing, yet nothing is just as meaningful. For adults, perhaps this is an irritating feature of modern life, but for our children it’s the fabric from which they are now being shaped.

They have no memory of the world before this flood. We are constructing the environment in which the next generation will learn what thinking feels like, what being human and being humane means. Perhaps we should grab hold of the places where learning still happens as part of life — the kitchen, the street, the broken bicycle, the blank page — places where education happens without announcing itself as such.

The danger of this new Age is that its definition of progress assumes that enough efficiency will finally bring us to equilibrium, and that equilibrium will bring abundance. We spent centuries trying to make knowledge abundant and seemingly succeeded. We now need to ask what all this abundance is for — and give a plausible answer.

Tatiana Dias is a fellow at Tech Policy Press.

Brazil is just days away from a runoff election in which the Workers’ Party (PT) candidate, incumbent president Luiz Inácio Lula da Silva, will face far-right Liberal Party (PL) candidate Flávio Bolsonaro, a senator and the son of former President Jair Bolsonaro. The race is extremely tight.

In the first round, Bolsonaro surprised many by earning more votes than Lula, while his party elected the largest far-right caucus in congressional history.

The fierce second-round campaign is playing out across social media, especially on Meta’s platforms — the main public arena for political debate in Brazil. More than 70% of Brazilians rely on Instagram 

. But day after day, evidence mounts that Meta has failed to ensure a level playing field for both candidates.

Allegations of censorship, account bans, reach suppression, and algorithmic discrimination that appear to disproportionately affect Lula and his supporters are piling up. 

 a complaint with the Superior Electoral Court (TSE) on Saturday, demanding fair treatment online. Lula’s campaign 

 Meta to international observers and to the Organization of American States. The cases are also under 

scrutiny from the Federal Prosecutor’s Office

, which opened an investigation on Wednesday to examine Meta’s conduct during the elections.

 sent to the New York Times, Meta denied political interference and said that “it is unfortunate that minor technical issues are being spun into a political narrative. Nothing could be further from the truth.”

from United States President Donald Trump. Tariffs imposed by the US government, the 

 — have put external pressure on the country that is also reverberating across Latin America. In this geopolitical context, the role of Big Tech companies that are politically aligned with Trump has created an atmosphere of suspicion.

Concerning evidence has been mounting in recent weeks, and the nonprofit organization I cofounded and direct—

—has collected many examples of questionable content moderation decisions. Before the first round, such decisions 

, including Meta’s blocking of ads from the PT campaign’s official account. Now, there are also thousands of reports of restricted users, complaints about bias in Meta AI outputs, and a “technical glitch” that undermined the performance of Lula’s videos on Instagram, all favoring Bolsonaro’s PL campaign.

According to the PT campaign, the far-right candidate’s videos 

 there were technical errors in how video views were counted globally, while denying any political interference in Brazil. “We identified a technical problem affecting some video metrics. We are working to resolve this issue as quickly as possible,” a spokesperson said.

Another basic failure was the lack of safeguards in Meta AI’s image generation related to candidates. Anyone who asks the tool to generate an image or a T-shirt design with “Flávio” and “22” — the far-right party’s ballot number — gets the request fulfilled, using the candidate’s real photo. But when asked to generate an image of Lula, the AI refuses, saying it was unable to generate the image. The test 

 by millions of Brazilians using a wide range of prompts.

Meta AI does generate images of Lula for users outside of Brazil or when using a VPN, which suggests a geo-targeted block.

 by a columnist for UOL, the Meta AI system responded: “You are right to point that out — there should not be different treatment. What happened here was: I tried to generate both banners you requested with ‘lula 13’ and the automatic safety system for images blocked it. When you requested with ‘flavio 22,’ the same system allowed it and I generated it.” In response to an additional prompt, the tool offered that “this violates the core principle of Brazilian electoral law.”

Hours after the story went viral, Meta told the press it was taking additional measures to ensure that prompts using either of the candidate’s first names and ballot number would not produce images. However, Meta AI 

 of Flávio Bolsonaro with minimal encouragement, but not of Lula.

Lula’s campaign brought the case to the TSE. “Meta AI, the artificial intelligence engine that powers all Meta-managed platforms — Instagram, WhatsApp, and Facebook — is, in itself, biased and seeks, at all costs, to deliver results favorable to Flávio Bolsonaro, even when the same user insists on Lula,” 

At the same time, users are reporting an unprecedented wave of post removals, profile bans, and other restrictions. CTRL+Z launched a survey shortly after the runoff was set and has so far collected more than 

of accounts that were censored, blocked, or had features restricted or experienced sudden drops in reach. More than 90% are from Lula supporters. The cases reveal what appear to be serious moderation failures and false positives, including activist posts that are misclassified as child sexual abuse material or terrorism-related content.

Accounts belonging to social movements — including journalists’ unions; student organizations such as university student councils and LGBTQIA+ collectives, state-level chapters of the left-wing Landless Workers’ Movement (MST) — were banned without explanation and fan accounts for Madonna that had announced their support for Lula, some with over 100,000 followers, were taken down (some have since been restored). Accounts linked to the PT, Lula’s party, and the election campaign had their functionalities and reach restricted.

The majority of complaints, however, involve feature restrictions and sudden drops in reach, which have affected ordinary citizens as well as major influencers — in some cases, by up to 90%. Instagram profiles of journalist Leandro Demori, and of musician Baco Exu do Blues, both critics of Bolsonaro, are among the prominent personalities who saw dramatic declines in their metrics.

Regarding the CTRL+Z survey, which was used as the basis for the Federal Prosecutor’s Office to open its investigation following a request by a coalition of CTRL+Z, Article 19, Conectas, Sleeping Giants Brasil and others, Meta objected that it had not been given access to the full dataset and responded to a request for comment saying that “a deliberately selected sample does not reflect the scope and rigor of our content moderation work, which is aimed at ensuring equal treatment across our platforms for all. We operate in compliance with applicable law and electoral regulations.”

Superior Electoral Court faces accusations of bias

As evidence against Meta piles up, civil society is demanding action from the Superior Electoral Court. The court’s president is Kássio Nunes Marques, a justice appointed by Bolsonaro who is 

 in the Banco Master case, Brazil’s biggest corruption scandal in recent years, including with 

 that he participated in a prostitution ring run by a banker named Daniel Vorcaro, who is currently in jail.

Flávio Bolsonaro also faces allegations of links to Vorcaro. His potential election could impact the future of investigations, possibly benefiting those involved.

In the first round, Nunes Marques made no controversial rulings. In the runoff, however, he moved to suspend “Exército do Lula” (Lula’s Army), a pro-Lula support WhatsApp group created by federal congressman André Janones, also of the PT, that had gathered 3.3 million members in just a few days through coordinated campaign actions. 

, the channel promoted disinformation and instructed users to act as “bots.” The far-right candidate himself, however, also has a similar group with the same types of content, which 

remains active despite violating TSE rules.

Nunes Marques’ decision caused friction among the justices, 

 by the TSE’s inspector general, Ricardo Villas Bôas Cueva, on the grounds of protecting freedom of expression, provided that those responsible for the channel refrain from disseminating or reproducing illicit content.

But the war of rulings continued. Two days later, again at the request of Bolsonaro’s party, Nunes Marques overturned Cueva’s decision and 

 the WhatsApp channel for 48 hours, arguing that it failed to comply with Meta’s terms of service. Before the period expired, however, the 

 after meeting a registration requirement.

In this context, Justice Nunes Marques has been 

Bolsonaro’s campaign in his rulings, and may face a disqualification motion filed by the Brazilian Association of Jurists for Democracy. PT’s president, congressman Rui Falcão, 

 to the Supreme Federal Court on grounds of Nunes Marques’ partisan bias.

Latest

Podcast

Recent

Our content delivered to your inbox

Thank you!

Around the globe