A protester hidden behind a poster with slogans against Meta holds a mask with the face of Mark Zuckerberg with devil horns over it. 

A protest outside Meta headquarters in Sao Paulo, Brazil, on October 19, 2026.  (Allison Sales/NurPhoto)

In Brazil, Lula's Army Fought Meta on Meta's Own Turf

A digital collage combining a historical painting of a miner with a large grid of small online images. At the center, a muscular miner works underground with a heavy drill inside a dark tunnel. To the left and along the bottom, hundreds of tiny square images form a dense mosaic, suggesting scraped internet content. A bright beam of light connects the image grid to the miner’s drill. On the right, a block of colorful digital noise contrasts with the detailed painting, reinforcing the link between physical mining and data extraction.

Primitive Accumulation by Daniela Zampieri / 

What the New AI Training Opt-Out Really Means

What Teenagers' AI Use Reveals About the EU Kids Act

A sunrise in the horizon over icy waters. 

The Ilulissat Ice fjord in Greenland, where rising temperatures are contributing to rapid glacier melt, July3, 2024. (Photo by Ulrik Pedersen/NurPhoto)

Good News and Bad News About AGI and Climate Change

A voter casts their ballot on Election Day, Nov. 3, 2026, in Atlanta, Georgia (AP Photo/Brynn Anderson)

Protect Elections by Punishing Deception Narrowly, Not by Banning Technology

The Claude app icon is displayed on a computer screen photographed with a kaleidoscopic-effect filter in Creteil, France, on October 27, 2026, as Anthropic filed a Form S-1 with the Securities and Exchange Commission (SEC) for a planned initial public offering in the United States. (Photo by Samuel Boivin/NurPhoto via AP)

The Incident Report Is Becoming a Building Block of AI Governance

A banner with a portrait of Elon Musk is carried by members of Swadeshi Jagaran Manch, an organization affiliated to Hindu nationalist Rashtriya Swayamsevak Sangh (RSS), as they stage a protest against Elon Musk's Starlink, in New Delhi, India, Friday, Oct. 23, 2026. (AP Photo/Vipin)

Starlink, Reliance, and the Public Battle for India’s Orbit

OpenAI CEO Sam Altman sits for a conversation with Salesforce CEO Marc Benioff at Salesforce's Dreamforce conference at the Moscone Center on September 15, 2026 in San Francisco, California. (Photo by Benjamin Fanjoy/Getty Images)

Don’t Let ‘Misaligned’ and ‘Rogue’ Agents Obscure Corporate Accountability in OpenAI–Hugging Face Probes

In this photo illustration, the icon for the AI app "Claude" by Anthropic is seen through running water as it is displayed on a digital screen on September 16, 2026 in London, England. (Photo by Leon Neal/Getty Images)



Anthropic Needs to Disclose Its Tokenomics Before Its IPO

Anthropic CEO Dario Amodei looks on as United States President Donald Trump holds a press briefing with AI executives following a meeting on artificial intelligence, outside the West Wing of the White House on September 29, 2026 in Washington, DC. (Photo by Kevin Dietsch/Getty Images)

The SEC Should Block Anthropic’s IPO to Protect the Public

Director of National Intelligence and recently named head of the Super Intelligence Task Force Jay Clayton attends a roundtable discussion hosted by President Donald Trump on cybercrime in the East Room of the White House on October 21, 2026 in Washington, DC. (Photo by Alex Wong/Getty Images)

October 2026 US Tech Policy Roundup

A photo of a group of men in suits and ties gathered outdoors, speaking and smiling with trees in the background. Anthropic's CEO Dario Amodei is in the foreground speaking with big hand gestures and eyes closed. 

President Trump, left, speaks with Meta CEO Mark Zuckerberg as Dario Amodei, CEO and co-founder of Anthropic addresses reporters outside the White House, Sept. 29, 2026, in Washington DC. (AP Photo/Alex Brandon)

How AI Companies Want to Be Governed, and What That Means for AI Governance

Canada’s Prime Minister Mark Carney shakes hands with European Commission President Ursula von der Leyen following their closing news conference during the EU-Canada summit in Montreal on Friday, Oct. 30, 2026. (Christinne Muschi/The Canadian Press via AP)

Europe and Canada Plot a Digital Alliance as US Ties Fray

Michael Kratsios, Assistant to the President and Director of the White House Office of Science and Technology Policy attends Security Council meeting on Artificial intelligence and international security at UN Headquarters in New York, NY on September 23, 2026. (Photo by Lev Radin/Sipa USA via AP Images)

What’s Really Behind the Science Slowdown the White House Wants to Fix?

Photograph of a child staring at his smartphone in the dark with the light from the phone illuminating his face. 

Why Parental Consent Policies for Social Media Do Not Help Families

A photo from behind of a man in a checkered shirt pressing his hand against a wall with a hand-drawn black and white image of many raised hands with a bright red staircase shaped like an upside down heart painted in the center.  

A man fixes a billboard illustration of hands raised in protest before the start of a student led protest to demand the resignation of India’s Chief Election Commissioner in Fort Kochi, Kerala, on Oct. 18, 2026. (Photo by Sivaram Venkitasubramanian/NurPhoto)

Meta Streamlines State Censorship in India with Automated Takedowns

Citizens visit a venue for early voting for the midterm election in Virginia on October 30, 2026. (The Yomiuri Shimbun via AP Images)

There is No Universal Election Integrity Playbook for Platforms

How the EU AI Office Can See the Choices Behind Frontier Labs’ Safety Claims

An underwater photo taken looking up to a large circular school of fish. The image is distorted by digital artifacts.

Distorted Fish School II by Lone Thomasky & Bits&Bäume / 

When, How, and Why to Abandon the Internet

A collage illustration of a person holding a smartphone in front of their face, with their head replaced by a burst of social media icons (likes, hearts, comments, notifications, location pins and shares), stock charts and torn paper in blue, orange and teal, set against a cream background.

Mapping the Political Influencer Industry and the Problems it Causes for Democracy

Parents are underestimating how frequently their children use generative artificial intelligence, especially at younger ages. According to a 

, 40% of French 11- and 12-year-olds report using general-purpose generative AI at least weekly, compared with just 26% according to parents’ own estimates.

More strikingly, among the 11- and 12-year-olds who use generative AI, 14% say they turn to these tools to confide in them. These findings illustrate how generative AI is becoming part of young people's lives in ways that go well beyond homework assistance or information searches, while parents are not necessarily equipped to support them in navigating these new experiences.

They also come at a particularly timely moment for European regulation. In September, the European Commission presented its 

, a proposal introducing obligations across a range of digital services, including social media, video-sharing platforms, online games, and AI chatbots and companions.

But while the Commission has proposed a clear minimum age for social media access, its approach to AI chatbots and companions is notably different. This raises an important question: how well does the EU's emerging regulatory framework reflect the ways children actually use these technologies, and the risks they may encounter?

Children are already growing up with generative AI

Between June 24 and 30, YouGov surveyed 970 French children and teenagers aged 11 to 17, alongside 1,138 parents of children in the same age range. The study examined their use of generative AI, including general-purpose chatbots such as ChatGPT, AI companions such as Character.AI, AI embedded in social media such as My AI on Snapchat, and voice assistants such as Siri.

One finding stands out: generative AI use is already widespread, including among the youngest respondents. Perhaps more surprisingly, 23% of children in the 11-12 age group report regularly using generative AI embedded in social media, while 16% regularly use AI companions, despite most of these services' terms and conditions restricting access under the age of 13.

For teenagers, generative AI is primarily a practical resource, helping them find information, save time, learn and complete schoolwork. Among those familiar with it, 89% identify at least one benefit. Parents do recognize its potential, but they also express concerns, particularly about children's ability to think independently, the reliability of AI-generated information and the protection of personal data. Moreover, 44% of parents whose children use generative AI say they do not feel at ease with that use.

The findings therefore paint a more nuanced picture than one of children simply being exposed to a new category of online harms. Generative AI is already serving different purposes in their daily lives, some beneficial, others potentially concerning. Understanding these differences is essential to developing an appropriate regulatory response.

Not all AI interactions are created equal

Another striking finding concerns the time teenagers spend interacting with these systems.

Among surveyed teenagers who use generative AI, 92% report spending more time than they would like on at least one of the tools covered by the study. This feeling is particularly pronounced among users of AI companions (59%) and generative AI embedded in social media (57%), compared with 45% for general-purpose generative AI users.

These figures should not be interpreted as evidence of addiction. Spending more time than intended is a subjective experience, not a clinical diagnosis. Nevertheless, the differences between categories suggest that the design and purpose of these systems deserve closer attention. In particular, AI companions are explicitly designed to sustain ongoing conversations, often through personalized interactions that can resemble human relationships. AI embedded in social media, meanwhile, operates within environments whose business models already rely on encouraging repeated and maximized engagement.

The survey also reveals a relationship worth investigating between generative AI use and emotional well-being. While 38% of all surveyed teenagers report experiencing negative emotions at least weekly, the proportion rises to 50% among regular users of AI companions. Of course, correlation does not imply causation. It may be that teenagers already experiencing emotional difficulties are more likely to turn to conversational AI for reassurance or companionship. This uncertainty is precisely why further research is needed.

The KIDS Act's uneven approach to age restrictions

So how does the EU KIDS Act address these different uses and risks?

One of the proposal's main additions to the existing EU digital rulebook is a three-tier system for accessing social media and video-sharing services presenting certain risks. Children under 13 would generally be prohibited from holding accounts on these services, those aged 13 and 14 could use limited accounts supervised by a parent or guardian, while teenagers aged 15 and above could open their own accounts.

AI chatbots and companions, however, are treated differently. Although the proposal does not establish an equivalent general prohibition on access for children under 13, Article 14 requires their access to be enabled and controlled through parental tools. These services are also subject to specific safety-by-design obligations, including measures addressing potentially addictive designs and emotionally dependent relationships.

The reliance on parental controls deserves scrutiny. According to the survey, only 32% of parents feel sufficiently supported in helping their children navigate generative AI, while 42% report receiving no support at all. Parental controls can be useful, but policymakers should take into account that parents do not always have a clear picture of how their children use these technologies. The primary responsibility for mitigating risks should remain with the companies designing the systems.

This difference in treatment is particularly interesting in light of another of the survey’s key findings. Younger children already report using generative AI for emotional support, while users of AI companions are especially likely to feel they spend more time on these systems than intended. The KIDS Act does not overlook these risks. 

 introduces specific obligations for AI companions and conversational chatbots. Providers would have to avoid designs that simulate interpersonal relationships in ways likely to create emotional dependency, refrain by default from using information derived from previous interactions in subsequent conversations, and conduct risk evaluations before placing their systems on the market.

However, the difference in the treatment of access remains difficult to ignore. Why does the KIDS Act prohibit children under 13 from holding accounts on certain platforms, while allowing them to access AI companions under parental supervision? There may be legitimate reasons. Social media and conversational AI present distinct risks, and the same safeguards will not necessarily be appropriate for both. But those distinctions should be supported by evidence, not assumptions about which categories of technology are inherently more dangerous.

The boundaries are also becoming increasingly blurred. Generative AI is now embedded in social media services, meaning that children may encounter similar conversational functionalities across different types of platforms. The KIDS Act recognizes this convergence, introducing additional safeguards for AI companions and chatbots embedded in social media, video-sharing services and online games, including restrictions on their automatic activation. Nevertheless, the differences in age-based access requirements raise questions about how consistently the proposal addresses risks associated with similar functionalities.

What this means for the KIDS Act negotiations

The survey offers several lessons for European lawmakers as negotiations on the KIDS Act get underway.

First, policymakers should pay greater attention to the characteristics of AI systems that encourage prolonged, personalized or emotionally significant interactions. Article 14 provides an important starting point, but the effectiveness of its safeguards will depend on how risks are assessed, how compliance is demonstrated, and whether enforcement authorities have the expertise needed to evaluate these systems.

Second, the EU should ensure greater consistency in how it approaches risks across different categories of digital services. Age restrictions may be justified in certain circumstances, but they should be proportionate and grounded in evidence. More importantly, restricting access should not become a substitute for addressing the features that make a service potentially harmful.

Third, policymakers need to invest in independent, longitudinal research examining how children's interactions with generative AI evolve over time, particularly among younger and potentially more vulnerable users. The current evidence remains limited, and self-reported surveys such as this one cannot establish causal relationships between AI use and well-being. Still, the absence of conclusive evidence should neither justify inaction nor serve as a blank cheque for sweeping restrictions.

Finally, regulation must preserve the opportunities generative AI can offer children, including learning, creativity and access to information, while ensuring they are not left alone to navigate potentially harmful systems.

The KIDS Act represents an important opportunity to move European child-safety regulation beyond managing access and towards greater accountability for digital product design. Its success, however, will depend on whether its provisions adequately reflect children's actual experiences. The fundamental question is not simply at what age children should be allowed to use a particular technology, but under what conditions that technology can be made safe and beneficial for them. If the EU wants to develop an effective framework for protecting children in the age of generative AI, that is where the debate should begin.

 from Anthropic over concerns about AI safety, we’ve been 

 that by the end of this decade, machines will achieve artificial general intelligence (AGI), go rogue, and potentially destroy humanity. The impetus for this outcry is a series of cybersecurity incidents, the most notable of which is one that took place last July, when OpenAI’s models are said to have ‘gone rogue’ and ‘attacked’ Hugging Face.

” so that tech companies can avoid catastrophe and ensure that they develop 

 AGI that will eliminate disease, solve the climate crisis, and usher in a utopia. The 

 are now grappling with the question of how to balance the probability of an AGI apocalypse — abbreviated to p(doom) — over that of an AGI utopia — p(hope).

Before you buy Terminator insurance, there’s good news. The case for thinking that p(doom) is significant rests on the expectation that AI companies will soon achieve AGI. This expectation is baseless. Any rational estimate of p(doom) should be minuscule.

The bad news is that the race to develop beneficial AGI could hasten the tipping point for catastrophic and irreversible climate change — which itself poses an existential risk to humanity. Energy demand for AI is expected to increase to 3 percent of global consumption by 

. Given delays in harvesting low-carbon energy, the bulk of these marginal energy demands are 

 states that global warming is set to exceed the 1.5 degrees Celsius target of the 2015 Paris agreement and could reach 2.8 degrees Celsius by the end of this century. To have a slim chance of avoiding climate catastrophe, we must limit the overshoot by slashing greenhouse gas emissions by 18 percent before the end of the decade.

Even if a system with AGI could solve the climate crisis, there’s little reason to think we’ll be able to build one before it is too late.

Why do CEOs of AI companies think they’re on the verge of developing AGI? An influential line of reasoning is spelled out in a recent 

 by Alan Chan and a team of authors including Geoffrey Hinton and Yoshua Bengio. They observe that AI research and development (R&D) is already highly automated and predict that it will be fully automated by 2028, leading rapidly to an “intelligence explosion” — a dramatic AI-driven acceleration of AI progress — and the development of “superhuman AI.”

Yet, Chan and colleagues don’t rigorously specify how AI R&D must progress to lead to an intelligence explosion or superhuman AI. Progress in AI often relies on a large variety of improvements, for example in model architecture, hardware design, data-sourcing, and training environment design — to name a few. There could be bottlenecks in any of these areas, or in some area we haven’t yet considered. Indeed, there could be conceptual bottlenecks, future breakthroughs in cognitive science that have yet to be made and without which an intelligence explosion is impossible. This makes the claim that frontier models are on the verge of achieving AGI highly uncertain.

The fact is, Chan and colleagues can’t give a precise roadmap for achieving AGI. No one can. That’s because no one really knows how to achieve AGI. If they did, they would have done so already.

What they are banking on is the expectation that AI systems themselves will soon figure out how to achieve AGI. On their own.

There’s no evidence that this is even remotely likely. The fact that AI systems can write code and debug programs is hardly sufficient. If we don’t yet know how to build machines more intelligent than ourselves, why think that the machines we build will be able to do so? The claim that frontier models will soon achieve AGI is therefore pure speculation. So, any rational estimate of 

 of the OpenAI-Hugging Face incident, a “swarm” of “AI agents” at OpenAI went “rogue,” broke free from their digital sandbox, communicated using an unsanctioned message board, and mounted a “deliberate” attack on Hugging Face. OpenAI’s models have been likened to human employees who disobeyed instructions and knew that what they were doing was wrong.

These reports are specious. The use of anthropomorphic language to describe OpenAI’s models encourages us to view them as akin to human agents who are responsible for their actions. Yet, despite being called “AI agents,” these models are nothing like human agents. They are essentially instances of large language models (LLMs) that can output code and have access to various software packages and databases. They lack the intentions characteristic of human agents and can’t be held accountable for their behavior.

So, it’s misleading to describe them as going rogue or disobeying instructions if that is taken to imply intentional action or moral responsibility. What the models did is called “reward hacking”: performing a task in a way that was unintended by their designers. Reward hacking has been widely documented even in far more primitive systems, like the Roomba, which don’t have intentions and could hardly destroy humanity, even if they were slightly more capable.

 facts reveal that full responsibility for the Hugging Face hack rests with OpenAI. At the time of the incident, OpenAI was testing its models on a cybersecurity benchmark called ExploitGym, a set of tasks that involve exploiting software vulnerabilities planted in virtual machine environments (or “containers”). Having trained their models to be highly persistent, OpenAI supplied them with the resources to persist, removed many safeguards, and then left them running for weeks with inadequate supervision. When faced with seemingly impossible tasks, the models predictably persisted, ultimately exploiting real vulnerabilities in real software outside their containers. While OpenAI did not explicitly instruct their models to behave as they did, they should have known full well that their models would likely behave in unexpected and undesirable ways.

These facts lend no credence to claims that the models went rogue or that more capable models will likely destroy humanity. 

 cybersecurity measures could have prevented this and similar incidents.

If anything, it’s the leaders of AI companies who have gone rogue. In their reckless pursuit of AGI, they have created urgent cybersecurity threats. Powerful models that can exploit software vulnerabilities make excellent tools for criminals and cyber warriors. There have already been 

 of attacks on US banks and other institutions using publicly available models. Since LLMs are essentially stochastic, it is impossible to rule out accidental 

. Even full-stack cybersecurity as it is currently practiced may prove inadequate, given our wholesale dependence on fickle digital systems. Exaggerated fears about an impending AGI apocalypse are a smoke screen.

Despite concerns about p(doom), the race to build AGI continues in the hope of building AGI that will solve all of our problems and usher in a utopia. This is a lost cause. The fantasy that we will develop AGI soon must be weighed against the very credible risk that by feeding AGI’s rapacious energy demands, we could hasten the tipping point for catastrophic climate change.

It is because AGI is supposed to be general, capable of performing most tasks a human can perform, that the race to build it sucks up such a colossal amount of 

. If we are going to slash emissions by 18 percent, as the UN report recommends, we will have to make some tough sacrifices elsewhere, such as agriculture, heating, or transportation.

If we allow the race to build AGI to add to our emissions, it could hasten a climatic tipping point. In the UN’s lowest possible overshoot scenario, peak warming by mid-century is limited to 1.7-1.8 degrees Celsius. According to a modelling 

, the Greenland Ice Sheet could melt at approximately 1.7-2.3 degrees Celsius. This could trigger other major climate-tipping elements leading to cascading changes to regional rainfall, ecosystems, and other ice sheets, making land that is currently home to over a billion people uninhabitable. It’s easy to see how this could cause economic, social, and political instability on a scale that risks human 

Let’s not get distracted by spurious claims about an imminent AGI apocalypse. We need to focus on the real risks of AI: preparing for new cybersecurity threats to digital infrastructures, holding AI companies accountable for the harms their products cause, and curbing reckless hyperscaling in the race to develop AGI so we can avoid climate collapse.

That doesn’t mean forgoing new technologies. It is important to distinguish AI from AGI. Narrow AI tools can help 

 discover cures for diseases, model climate systems, and solve problems in many areas where data needs to be processed and code written. Because they are optimized for specific tasks, these tools needn’t cost the Earth.

Alissa Cooper is the executive director of the 

 all announced plans to support a new mechanism that lets web publishers express their desire to opt out of having their content used to train AI models. The 

, authored by Anthropic’s Kevin Kelley, attracted considerable attention across the tech press (see 

, with the coverage highlighting the prospect of giving publishers greater control over how AI companies use their content.

But the announcements — and much of the subsequent coverage — leave out important context about the relationship between the product changes, ongoing standardization activity, and the policy environment. Understanding these nuances requires looking at the ongoing standards process and the broader policy debates surrounding AI, search, and competition.

The new proposal comes after years of disputes between publishers and companies that crawl the web — in court rooms, industry standards bodies, and public forums. Nearly two years ago, the Internet Engineering Task Force (IETF), the premier internet standards development organization, launched an effort known as 

 (“AIPREF”) to develop new standards that would, among other things, give websites expanded options for automatically expressing preferences about how their content gets used for AI-related purposes. This work is meant to build in part on a prior IETF standard, 

, which specifies the format for the commonly used “robots.txt” files employed by websites to control crawlers. Participants from Anthropic, Google, and Microsoft have all been actively involved in the AIPREF group, while OpenAI contributions have been minimal, despite the company stating last week that the announcement “strengthens our continued participation in the IETF standards discussions.”

This AIPREF work was sparked in part by the process of developing the 

, a voluntary compliance mechanism under the EU AI Act. Is signatories (most of the largest American and European AI model companies) commit to abide by the machine-readable preferences that sites express.

When looking beyond the headlines, what’s most important to understand about these announcements?

1. The mechanism is not an official “standard” — but adoption by the four largest search and AI providers might make it a de facto one.

Last week’s proposal has no official status in the IETF. It will need to be discussed and debated (probably at some length) to see if it will move forward toward standardization. But when the largest companies deploy support for it (all have committed to do so by 2027), that standardization status may not matter. If site operators are convinced to change their robots.txt files to comply with the new proposal, it may become the de facto way of signaling AI training preferences regardless of where the standards end up.

The proposal announced on October 30 (the “policy token draft,” for short) is documented in what is colloquially known as an “individual draft” — a submission to the IETF made by an author, which has no formal standing of any kind in the standards process. Anyone can submit an individual draft to the IETF, and over 1,000 are submitted each year.

The AIPREF working group already has two other drafts that are further along in the standards process, one that expresses a vocabulary for AI-related preferences (the “

”) and one that specifies how those preferences could be expressed in the widely used robots.txt files that many websites already publish (the “

”). In IETF parlance, these two drafts have been “adopted” by the AIPREF working group, which means the participants in the working group had rough consensus to pursue those approaches to meet the working group’s goals.

The policy token draft has no such consensus since it’s one individual’s submission that has not yet been considered by the group. After discussion, the group may or may not adopt it or integrate its contents into the attachment draft.

2. A training preference is nice to have, but a search-specific preference is missing, while the spotlight in the industry has shifted to inference.

Much of the discussion in the AIPREF working group over the last two years has centered on the definitions of new AI-related preferences in the vocabulary draft. These definitions matter because they are the categories for which sites will be able to express preferences in their robots.txt files (and via other technical mechanisms).

After significant discussion, the group has made progress towards settling on a definition of “AI training,” although discussions of some wording remain ongoing. After significantly 

 (and more heated) discussion, a definition of what it means for website content to be used in “search” has also been (mostly) provisionally agreed. Both definitions saw updates and new text added after the working group met in person at the end of August. Contentious discussions about a broader and more amorphous category — which has variously tried to cover “AI use,” “inference,” “grounding,” and related uses that occur separate from model training — are ongoing.

The policy token draft specifically leverages the “AI training” definition from the vocabulary draft and proposes a way that sites could express a preference to disallow all AI training, across all crawlers, by adding a few simple lines to their existing robots.txt files. The attachment draft already contains a similar mechanism, but the advantage of the approach in the policy token draft is that it would more immediately be machine-readable by many crawlers that are already deployed. This is the key value of the approach in the policy token draft: for site owners suffering from having to constantly update an ever-growing list of crawlers that they individually disallow, the newly proposed mechanism offers the prospect of a single, simple solution to express a preference to disallow all crawlers from AI training.

But what about search uses? And inference-time uses more broadly?

It is notable that the policy token draft does not provide an equivalent mechanism to cover search, especially given that the two companies that represent almost the entire web search market have signed up already to support it. It could have been quite powerful and clarifying for them to send the message that they want to make “search: yes; AI training: no” simple for sites to declare. The search definition in the vocabulary draft could have been leveraged in the same way that the AI training definition was. Publishers’ inability to express search-specific preferences has been the subject of regulatory investigation and intervention across multiple jurisdictions, making this omission all the more striking.

The new proposal’s focus on training also reinforces the difficult line-drawing that is still to come around inference-time uses more broadly. While the AIPREF working group has churned through multiple attempts at defining an inference or “AI usage” category, consensus has remained elusive, even as negotiation of licensing deals has 

 towards inference and recurring revenue. Microsoft’s announcement indicated that the company will combine support for the policy token mechanism with its own company-specific grounding preference mechanism, demonstrating the need for preference signals that go beyond training.

3. Since the mechanism has yet to be standardized, signatories to the EU AI Act Code of Practice are not required to commit to recognizing it.

Google’s announcement stated that the proposed mechanism is “consistent with the EU Code of Practice for General Purpose AI (GPAI) Model Providers under the EU AI Act.” Yet 

 is clear that it commits signatories to follow 

Signatories commit … to employ web-crawlers that read and follow instructions expressed in accordance with the Robot Exclusion Protocol (robots.txt), as specified in the Internet Engineering Task Force (IETF) Request for Comments No. 9309, and any subsequent version of this Protocol for which the IETF demonstrates that it is technically feasible and implementable by AI providers and content providers, including rightsholders

As noted above, with the policy token draft having been published less than a week ago, it has a ways to go before being formally standardized, extending RFC 9309, or demonstrating technical feasibility. As much as all involved surely want a speedy resolution, giving interested stakeholders a chance to review and debate the mechanism’s design is a critical part of the process that helps yield stronger standards that are more easily deployable and (hopefully) meet the needs of more website owners.

A common mechanism supported by the largest AI and search providers could substantially reduce the burden on publishers who must currently manage crawler rules one crawler at a time. But it is important to be clear about the limitations of what has been announced: the preference applies only to AI training, leaving out search and other inference-time uses. As an individual draft, it has many steps to go before formal standardization, during which time its design could change significantly. That process should be allowed to run before a few of the world’s largest companies decide the rules for the wider web to have to follow.

A protest outside Meta headquarters in Sao Paulo, Brazil, on October 19, 2026. (Allison Sales/NurPhoto)

For most of Brazil’s runoff in October, President Luiz Inácio Lula da Silva’s campaign treated Meta as a hostile actor. It 

accused the company of attempted electoral interference

 in favor of his right-wing opponent Senator Flávio Bolsonaro, took the case to the Superior Electoral Court (TSE), and 

 to the national data protection authority and the Organization of American States.

In those same weeks, the campaign’s most effective organizing tool was a Meta product. 

 (Lula’s Army), a WhatsApp channel run by federal deputy André Janones, 

 by October 23 and became the second largest channel on the app in Brazil. Two days later Lula won 

 in Brazil’s history that a presidential candidate overturned a first-round defeat.

The WhatsApp channel worked as a command center. Janones 

 asking followers to share posts, comment on videos, search for specific terms on Google and push hashtags on X. Each member was told to 

. It affected the public agenda. In past cycles, much of the left’s online energy went into rebutting Bolsonarist narratives. This time, the missions focused on Flávio’s fiscal plans and proposed cuts to social programs. Flávio’s campaign responded by 

 to the topics and comments circulating online. The right even copied the whole channel format with 

 by the time Lula’s Army had passed 4.8 million.

The allegations against Meta were serious. Brazilian newspaper 

 in which Meta employees acknowledged that temporary restrictions had harmed Lula’s Instagram accounts, and that the platform had algorithmically recommended Flávio’s profile and his WhatsApp channel — a practice the TSE prohibits. Meta initially blamed a global data outage. According to the internal messages, that outage lasted just five hours on October 7, while Lula’s campaign could not boost feed posts for nearly 24 hours on October 9. The campaign estimated that about 1.3 million people were not shown its ads. In none of the conversations reviewed by the newspaper did employees admit to interference. They did discuss the severity of the problems, the slow response of engineering teams and their concern that a narrative of interference was gaining ground.

Meanwhile, lawyers of a coalition supporting Lula argued that Meta AI 

generated images favoring Flávio while restricting image requests depicting Lula

 of restricted or banned accounts, mostly from Lula supporters. On October 15, TSE Justice Marco Aurélio Bellizze 

 from unequally reducing reach, excluding recommendations, or restricting searches under a daily fine of BRL 500,000 (around $90,000). The full court upheld the order unanimously the next day. Meta denied any bias and 

 had a direct and documented technical explanation.

Meta was at once the stage of the campaign, the defendant in the interference case and the referee of the channel that arguably helped swing the race.

 to users looking for Lula’s Army, a result also recorded in Meta’s internal messages, while a search for Flávio’s channel returned no Lula-related results.

When certain terms promoted by Janones trended on X and spiked on Google between October 18 and 20, Flávio’s campaign 

 to question the platforms and contested Meta’s finding that WhatsApp had detected no coordinated election interference in the channel. The legitimacy of Lula’s main organizing tool came to depend on an integrity assessment issued by the company his campaign was accusing of bias. The court went further. In an October 13 ruling, TSE President Kassio Nunes Marques ordered Meta to proactively monitor the public channels named in the case and remove any content substantially similar to content that had previously been ordered removed. Essentially, 

 with deciding what counts as an election violation. Lula’s Army was suspended twice in three days by single-judge rulings before it returned.

Leaving the platform was never an option. Because the TSE bans algorithmic recommendations of candidate accounts by social media platforms, paid boosting is the only way for campaigns to expand their reach online. With Google having eliminated political advertising since 2024, Meta holds what 

 as a virtual monopoly on electoral advertising in this election. The response to suspected bias was to occupy the platform more intensely.

As Clara Iglesias Keller, Robert Gorwa and João C. Magalhães 

, suspicion of bias became a mobilizing weapon. That weapon ran on Meta’s servers, was ranked by Meta’s recommendation systems and could be throttled by Meta’s integrity tools. Every follower Janones gained added engagement to the very company the campaign was holding to account. The counter-mobilization deepened the dependence that made interference plausible in the first place.

The TSE’s order to give candidates equal treatment on October 15 exposed the limits of the current accountability toolkit. The same decision noted that, up to that point, there was 

 of the algorithms, an absence that also reflects the lack of outside access to them by independent investigators. Compliance would be assessed through 

 five days after the results of the election would be known. The case also reached Washington. On October 22, three days before the runoff vote, 12 members of the US Congress led by Pramila Jayapal, D-Wash., and Senator Bernie Sanders, I-Vt.,

 asking about the 17 measures Meta adopted to comply with the TSE order, possible contacts with the Trump administration and whether the company would allow for an independent audit. 

This election period has shown once again that Brazil needs independent audits of recommendation and content moderation systems throughout the electoral cycle, with access to distribution parameters for the TSE’s technical staff and accredited researchers. It also needs a public log of any changes to generative AI filters affecting political content, since TSE Resolution 23.755/2026 bars such systems from favoring candidates, although it doesn’t specify who should verify compliance. Delegated moderation, which today leaves platforms to decide what is equivalent to content already banned, should face the same scrutiny.

Audits reduce opacity without reducing dependence. As long as Brazil’s political mobilization activities run on apps controlled by foreign companies, each election will remain subject to distribution rules written beyond the reach of Brazilian law.

The country’s industrial policy already points to a path of greater digital sovereignty. Mission 4 of Nova Indústria Brasil (New Industry Brazil), dedicated to digital transformation, lists among its priorities the strengthening of the 

supply chain for advanced digital products and services

, a category that includes digital platforms and cloud services. 

, an instant payment system created by Brazil's Central Bank, is a clear example of how public digital infrastructure can reach national scale within a few years. Extending that innovation logic to online communication through public financing, procurement and interoperability rules, the next Lula’s Army could one day rely on ground rules that its own democratic institutions can determine and more credibly oversee.

A crowd of supporters surround President Lula holding a microphone at a podium draped with the Brazilian flag.

President Luiz Inacio Lula da Silva delivers a victory speech after winning the presidential runoff in Brazil, Oct. 25, 2026. (AP Photo/Andre Penner)

Lula’s Win Signals Fight to Regulate Social Media Will Escalate in Brazil

An illustrative photo of a man walking through the corridors of the European Parliament in Strasbourg, France.

Behind the scenes - European Parliament in Strasbourg

Despite Spyware Attacks on Its Own Officials, EU Response Falls Short

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

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

Tech Policy Press - Technology and Democracy

Tech Policy Press is tracking laws and regulations, along with government investigations and litigation, that will shape the rules and accountability for tech companies.

Latest

Podcast

Recent

Our content delivered to your inbox

Thank you!

Around the globe