LONDON — The U.K.’s data protection watchdog has opened a formal investigation into Elon Musk’s companies X and xAI, over the use of personal data by the Grok AI system to generate a flood of sexualized deepfakes. In a statement on Tuesday, the Information Commissioner’s Office said the “reported creation and circulation of such content raises serious concerns under U.K. data protection law and presents a risk of significant potential harm to the public.” “These concerns relate to whether personal data has been processed lawfully, fairly and transparently, and whether appropriate safeguards were built into Grok’s design and deployment to prevent the generation of harmful manipulated images using personal data,” it said. The formal investigation follows an announcement last month that the ICO was seeking urgent information from X and xAI, amid widespread reports that Grok had been used to generate sexualized images of children and adults. William Malcolm, executive director for regulatory risk and innovation at the ICO, said the reports about Grok “raise deeply troubling questions about how people’s personal data has been used.” “Losing control of personal data in this way can cause immediate and significant harm. This is particularly the case where children are involved,” Malcolm said. “Where we find obligations have not been met, we will take action to protect the public.” While the ICO’s investigation will focus on X and xAI’s compliance with U.K. data protection law, Malcolm said it would work closely with other regulators in the U.K. and abroad that are also investigating the issue. Ofcom, the U.K.’s communications regulator, opened a formal investigation into X last month under the Online Safety Act. That investigation is ongoing, Ofcom said on Tuesday. It is progressing “as a matter of urgency” but could take “months,” Ofcom added, noting that it must follow a “fair process” and “it would not be appropriate to provide a running commentary.” Ofcom also said it is not currently investigating xAI, which provides the standalone Grok AI tool, noting that “it can only take action on online harms covered by the [OSA].” The act does not apply to AI tools which do not involve searching the internet, interacting with other social media users, or generating pornography, it said. The U.K.’s Technology Secretary Liz Kendall has previously said she is assessing options to address “gaps” in the OSA. The European Commission announced its own probe into X last month, while French authorities searched X’s offices in Paris on Tuesday as part of their own criminal investigation into Grok, POLITICO reported. X did not immediately respond when contacted for comment.

BRUSSELS — The European Union is pressing ahead with talks to grant United States border authorities unprecedented access to Europeans’ data, despite growing concerns about American surveillance. The European Commission is brokering a deal to exchange information about travelers, including fingerprints and law enforcement records, so the U.S. can determine if they “pose a risk to public security or public order,” according to official documents. Commission officials flew to Washington last week for the first round of negotiations, according to two people familiar with the matter. The Trump administration’s request for deeper access comes after the U.S. border agency in December proposed reviewing five years of social media history. Talks are happening as the U.S. Immigration and Customs Enforcement (ICE) service is under heavy scrutiny for its use of surveillance technology against protesters in cities such as Minneapolis. The negotiations should be “put on hold” until the security and privacy of citizens in the EU and U.S. can be guaranteed, liberal European Parliament member Raquel García Hermida-van der Walle said in an interview. Romain Lanneau, a legal researcher with surveillance watchdog Statewatch, said police databases in Europe could contain information on anyone from protesters to journalists who might be considered a “threat,” and that — under the deal being discussed — this information would be at the fingertips of U.S. border authorities who could refuse those people entry to the United States or even detain them. European regulators are “very cautiously looking at what’s happening in the United States,” Wojciech Wiewiórowski, the EU’s in-house data protection supervisor, told POLITICO. Europe “has to be careful” about how it allows the data of Europeans to flow to the U.S., he said.  Hermida-van der Walle in January co-signed a letter by six prominent lawmakers calling on the Commission to stand down given the “current geopolitical context,” despite Washington’s admonition that failure to reach a deal will mean Europeans lose access to its visa waiver program. UNPRECEDENTED ACCESS The U.S. is seeking access to information including biometric data such as fingerprints that is stored on national databases in European countries, according to an explanatory note sent to national experts. The data would be used to “address irregular migration and to prevent, detect, and combat serious crime and terrorist offences,” the note said. In an earlier opinion on the deal, the European Data Protection Supervisor (EDPS) — a watchdog that advises the Commission on privacy policies — noted the deal would be the first of its kind to enable “large-scale sharing of personal data … for the purpose of border and immigration control” with a non-EU country. The Commission would negotiate a framework deal that would serve as a template for bilateral agreements called Enhanced Border Security Partnerships (EBSPs), which national governments agree with Washington. EU countries in December signed off on the Commission’s request to start talks with the U.S. Washington is pressuring its EU counterparts by imposing a deadline for the bilateral deals to be agreed by the end of 2026. If countries fail to reach a deal with the U.S. they risk being cut from the latter’s visa waiver program. The U.S has made it mandatory for all countries that are part of the visa waiver program to have an EBSP in place. “The pressure which the United States is extorting on our member states, the threats that if you don’t agree with this we will cancel your access to the visa waiver program, that is an element of blackmail that we cannot let go,” Hermida-van der Walle said. The EDPS watchdog has cautioned that the scope of data sharing should be as narrow as possible, with clear justifications for every query; transparency around how the data is used; and judicial redress available in the U.S. for any person. Commission spokesperson Markus Lammert emphasised at a recent press briefing that the framework being negotiated will involve “clear and robust safeguards on data protection,” and will ensure “a non-systematic nature of the information exchange and that the exchange is limited to what is strictly necessary to achieve the objectives of this cooperation.”  US PRIVACY UNDER PRESSURE Access to the data is the latest issue putting pressure on a troubled relationship between the U.S. and the EU on data privacy. Since whistleblower Edward Snowden in 2013 revealed U.S. mass surveillance practices affecting Europeans, the EU has tightened controls on how Washington handles Europeans’ data. Since the return of Donald Trump as president last year, officials and rights groups have deplored a move by the U.S. administration to gut a key privacy watchdog tasked with overseeing privacy safeguards in place to protect Europeans. The Trump administration has also been ramping up mass surveillance of citizens by federal agencies like ICE, including through contracts with Israeli spyware company Paragon, surveillance giant Palantir and other firms. Capgemini, a prominent French IT firm, on Sunday said it was selling off its American activities after it faced political backlash from the French government that its software was being used by ICE authorities. Civil rights groups, lawmakers and other watchdogs fear the new EU-U.S. data sharing deals would add to backsliding on privacy rights.    “The current initiatives are being presented as toward counter-terrorism, but a lot of them are actually adopted for the chilling effect [on political activism],” Statewatch’s Lanneau said. Hermida-van der Walle, the liberal lawmaker, warned: “If people have to go to the United States, if it’s not a choice but something that they have do, there is a risk of self-censoring.”  “This comes from an administration who claims to be the biggest defender of free speech. What they’re doing with their actions is curtailing the possibility of people to express themselves freely, because otherwise they might not get access into the country,” she said.

China’s foreign ministry on Wednesday said a new European Commission proposal to restrict high-risk tech vendors from critical supply chains amounted to “blatant protectionism,” warning European officials that Beijing will take “necessary measures” to protect Chinese firms. Beijing has “serious concerns” over the bill, Chinese foreign ministry spokesperson Guo Jiakun told reporters, according to state news agencies’ reports. “Using non-technical standards to forcibly restrict or even prohibit companies from participating in the market, without any factual evidence, seriously violates market principles and fair competition rules,” Guo said. The European Commission on Tuesday unveiled its proposal to revamp the bloc’s Cybersecurity Act. The bill seeks to crack down on risky technology vendors in critical supply chains ranging across energy, transport, health care and other sectors. Though the legislation itself does not name any specific countries or companies, it is widely seen as being targeted at China. 5G suppliers Huawei and ZTE are in the EU’s immediate crosshairs, while other Chinese vendors are expected to be hit at a later stage. European Commission spokesperson Thomas Regnier responded to the Chinese foreign ministry, saying Europe has allowed high-risk vendors from outside the EU in strategic sectors for “far too long.” “We are indeed radically changing this. Because we cannot be naive anymore,” Regnier said in a statement. The exclusion of high-risk suppliers will always be based on “strong risk assessments” and in coordination with EU member countries, he said. China “urges the EU to avoid going further down the wrong path of protectionism,” the Chinese foreign ministry’s Guo told reporters. He added the EU bill would “not only fail to achieve so-called security but will also incur huge costs,” saying some restrictions on using Huawei had already “caused enormous economic losses” in Europe in past years. European telecom operators warned Tuesday that the law would impose multi-billion euro costs on the industry if restrictions on using Huawei and ZTE were to become mandatory across Europe. A Huawei spokesperson said in a statement that laws to block suppliers based on their country of origin violate the EU’s “basic legal principles of fairness, non-discrimination, and proportionality,” as well as its World Trade Organization obligations. The company “reserve[s] all rights to safeguard our legitimate interests,” the spokesperson said. ZTE did not respond to requests for comment on the EU’s plans.

A clash between Poland’s right-wing president and its centrist ruling coalition over the European Union’s flagship social media law is putting the country further at risk of multimillion euro fines from Brussels. President Karol Nawrocki is holding up a bill that would implement the EU’s Digital Services Act, a tech law that allows regulators to police how social media firms moderate content. Nawrocki, an ally of U.S. President Donald Trump, said in a statement that the law would “give control of content on the internet to officials subordinate to the government, not to independent courts.” The government coalition led by Prime Minister Donald Tusk, Nawrocki’s rival, warned this further exposed them to the risk of EU fines as high as €9.5 million. Deputy Digital Minister Dariusz Standerski said in a TV interview that, “since the president decided to veto this law, I’m assuming he is also willing to have these costs [of a potential fine] charged to the budget of the President’s Office.” Nawrocki’s refusal to sign the bill brings back bad memories of Warsaw’s years-long clash with Brussels over the rule of law, a conflict that began when Nawrocki’s Law and Justice party rose to power in 2015 and started reforming the country’s courts and regulators. The EU imposed €320 million in penalties on Poland from 2021-2023. Warsaw was already in a fight with the Commission over its slow implementation of the tech rulebook since 2024, when the EU executive put Poland on notice for delaying the law’s implementation and for not designating a responsible authority. In May last year Brussels took Warsaw to court over the issue. If the EU imposes new fines over the rollout of digital rules, it would “reignite debates reminiscent of the rule-of-law mechanism and frozen funds disputes,” said Jakub Szymik, founder of Warsaw-based non-profit watchdog group CEE Digital Democracy Watch. Failure to implement the tech law could in the long run even lead to fines and penalties accruing over time, as happened when Warsaw refused to reform its courts during the earlier rule of law crisis. The European Commission said in a statement that it “will not comment on national legislative procedures.” It added that “implementing the [Digital Services Act] into national law is essential to allow users in Poland to benefit from the same DSA rights.” “This is why we have an ongoing infringement procedure against Poland” for its “failure to designate and empower” a responsible authority, the statement said. Under the tech platforms law, countries were supposed to designate a national authority to oversee the rules by February 2024. Poland is the only EU country that hasn’t moved to at least formally agree on which regulator that should be. The European Commission is the chief regulator for a group of very large online platforms, including Elon Musk’s X, Meta’s Facebook and Instagram, Google’s YouTube, Chinese-owned TikTok and Shein and others. But national governments have the power to enforce the law on smaller platforms and certify third parties for dispute resolution, among other things. National laws allow users to exercise their rights to appeal to online platforms and challenge decisions. When blocking the bill last Friday, Nawrocki said a new version could be ready within two months. But that was “very unlikely … given that work on the current version has been ongoing for nearly two years and no concrete alternative has been presented” by the president, said Szymik, the NGO official. The Digital Services Act has become a flashpoint in the political fight between Brussels and Washington over how to police online platforms. The EU imposed its first-ever fine under the law on X in December, prompting the U.S. administration to sanction former EU Commissioner Thierry Breton and four other Europeans. Nawrocki last week likened the law to “the construction of the Ministry of Truth from George Orwell’s novel 1984,” a criticism that echoed claims by Trump and his top MAGA officials that the law censored conservatives and right-wingers. Bartosz Brzeziński contributed reporting.

WARSAW — Poland’s nationalist President Karol Nawrocki on Friday sided with his ally U.S. President Donald Trump to veto legislation on enforcing the EU’s social media law, which is hated by the American administration. Trump and his top MAGA officials condemn the EU’s Digital Services Act — which seeks to force big platforms like Elon Musk’s X, Facebook, Instagram to moderate content — as a form of “Orwellian” censorship against conservative and right-wingers. The presidential veto stops national regulators in Warsaw from implementing the DSA and sets Nawrocki up for a a clash with centrist pro-EU Prime Minister Donald Tusk. Tusk’s parliamentary majority passed the legislation introducing the DSA in Poland. Nawrocki argued that while the bill’s stated aim of protecting citizens — particularly minors — was legitimate, the Polish bill would grant excessive power to government officials over online content, resulting in “administrative censorship.”  “I want this to be stated clearly: a situation in which what is allowed on the internet is decided by an official subordinate to the government resembles the construction of the Ministry of Truth from George Orwell’s novel 1984,” Nawrocki said in a statement — echoing the U.S.’s stance on the law. Nawrocki also warned that allowing authorities to decide what constitutes truth or disinformation would erode freedom of expression “step by step.” He called for a revised draft that would protect children while ensuring that disputes over online speech are settled by independent courts. Deputy Prime Minister and Digital Affairs Minister Krzysztof Gawkowski dismissed Nawrocki’s position, accusing the president of undermining online safety and siding with digital platforms.  “The president has vetoed online safety,” Gawkowski told a press briefing Friday afternoon, arguing the law would have protected children from predators, families from disinformation and users from opaque algorithms.  The minister also rejected Nawrocki’s Orwellian comparisons, saying the bill explicitly relied on ordinary courts rather than officials to rule on online content. Gawkowski said Poland is now among the few EU countries without national legislation enabling effective enforcement of the DSA and pledged that the government would continue to pursue new rules. The clash comes as enforcement of the social media law has become a flashpoint in EU-U.S. relations.  Brussels has already fined Elon Musk’s X €120 million for breaching the law, prompting a furious response from Washington, including travel bans imposed by the Trump administration on former EU Commissioner Thierry Breton, an architect of the tech law, and four disinformation experts. The DSA allows fines of up to 6 percent of a company’s global revenue and, as a measure of last resort, temporary bans on platforms. Earlier this week, the European Commission expanded its investigation into X’s AI service Grok after it started posting a wave of non-consensual sexualized pictures of people in response to X users’ requests. The European Commission’s digital spokesperson Thomas Regnier said the EU executive would not comment on national legislative procedures. “Implementing the DSA into national law is essential to allow users in Poland to benefit from the same DSA rights, such as challenging platforms if their content is deleted or their account suspended,” he said. “This is why we have an ongoing infringement procedure against Poland. We have referred Poland to the Court of Justice of the EU for failure to designate and empower the Digital Services Coordinator,” in May 2025, Regnier added. Gawkowski said that the government would make a quick decision on what to do next with the vetoed bill but declined to offer specifics on what a new bill would look like were it to be submitted to parliament again. Tusk four-party coalition does not have enough votes in parliament to override Nawrocki’s vetoes. That has created a political deadlock over key legislation efforts by the government, which stands for reelection next year. Nawrocki, meanwhile, is aiming to help the Law and Justice (PiS) political party he’s aligned with to retake power after losing to Tusk in 2023. Mathieu Pollet contributed reporting.

The Dutch government has quietly removed Google tracking tools from job listings for its intelligence services over concerns that the data would expose aspirant spies to U.S. surveillance. The intervention would put an end to Google’s processing of the data of job seekers interested in applying to spy service jobs, after members of parliament in The Hague raised security concerns. The move comes at a moment when trust between the Netherlands and the United States is fraying. It reflects wider European unease — heightened by Donald Trump’s return to the White House — about American tech giants having access to some of their most sensitive government data. The heads of the AIVD and MIVD, the Netherlands’ civilian and military intelligence services, said in October that they were reviewing how to share information with American counterparts over political interference and human rights concerns. In the Netherlands, government vacancies are listed on a central online portal, which subsequently redirects applicants to specific institutions’ or agencies’ websites, including those of the security services. The government has now quietly pulled the plug on Google Analytics for intelligence-service postings, according to security expert Bert Hubert, who first raised the alarm about the trackers earlier this year. Hubert told POLITICO the job postings for intelligence services jobs no longer contained the same Google tracking technologies at least since November. The move was first reported by Follow the Money. The military intelligence service MIVD declined to comment. The interior ministry, which oversees the general intelligence service AIVD, did not respond to a request for comment at the time of publication. In a statement, Communications Manager for Google Mathilde Méchin said: “Businesses, not Google Analytics, own and control the data they collect and Google Analytics only processes it at their direction. This data can be deleted at any time.” “Any data sent to Google Analytics for measurement does not identify individuals, and we have strict policies against advertising based on sensitive information,” Méchin said. ‘FUTURE EMPLOYEES AT RISK’ Derk Boswijk, a center-right Dutch lawmaker, raised the alarm about the tracking of job applicants in parliamentary questions to the government in January. He said that while China and Russia have traditionally been viewed as the biggest security risks, it is unacceptable for any foreign government — allied or not — to have a view into Dutch intelligence recruitment. “I still see the U.S. as our most important ally,” Boswijk told POLITICO. “But to be honest, we’re seeing that the policies of the Trump administration and the European countries no longer necessarily align, and I think we should adapt accordingly.” The government told Boswijk in February it had enabled privacy settings on data gathered by Google. The government has yet to comment on Boswijk’s latest questions submitted in November. Hubert, the cybersecurity expert, said the concerns over tracking were justified. Even highly technical data like IP addresses, device fingerprints and browsing patterns can help foreign governments, including adversaries such as China, narrow down who might be seeking a job inside an intelligence agency, he said. “By leaking job applications so broadly, the Dutch intelligence agencies put their future employees at risk, while also harming their own interests,” said Hubert, adding it could discourage sought-after cybersecurity talent that agencies are desperate to attract. Hubert previously served on a watchdog committee overseeing intelligence agencies’ requests to use hacking tools, surveillance and wiretapping.  One open question raised by Dutch parliamentarians is how to gain control over the data that Google gathered on aspiring spies in past years. “I don’t know what happens with the data Google Analytics already has, that’s still a black box to me,” said Sarah El Boujdaini, a lawmaker for the centrist-liberal Democrats 66 party who oversees digital affairs. The episode is likely to add fuel to efforts to wean off U.S. technologies — which are taking place across Europe, as part of the bloc’s “technological sovereignty” drive. European Parliament members last month urged the institution to move away from U.S. tech services, in a letter to the president obtained by POLITICO. In the Netherlands, parliament members have urged public institutions to move away from digital infrastructure run by U.S. firms like Microsoft, over security concerns. “If we can’t even safeguard applications to our secret services, how do you think the rest is going?” Hubert asked. The country also hosts the International Criminal Court, where Chief Prosecutor Karim Khan previously lost access to his Microsoft-hosted email account after he was targeted with American sanctions over issuing an arrest warrant for Israeli Prime Minister Benjamin Netanyahu. The ICC in October confirmed to POLITICO it was moving away from using Microsoft Office applications to German-based openDesk.

BRUSSELS — You can even put an exact date on the day when Brussels finally gave up on its decade-long dream of seeking to be the predominant global tech regulator that would rein in American tech titans like Google and Apple.  It came last Wednesday — Nov. 19 — when the European Commission made an outright retreat on its data and privacy rules and hit pause on its AI regulation, all part of an attempt to make European industries more competitive in the global showdown with the United States and China.   It sounded the death knell for what has long been described as the “Brussels Effect” — the idea that the EU would be a trailblazer on tech legislation and set the world’s standards for privacy and AI.  Critics say Washington is now setting the deregulatory trajectory, while U.S. President Donald Trump is battering down Europe’s ambitions by threatening to roll out tariffs against countries that he accuses of attacking “our incredible American Tech Companies.” “I don’t hear anybody in Brussels saying ‘We’re a super regulator’ anymore,” said Marietje Schaake, who shaped Europe’s tech rulebooks as a former European Parliament member and special adviser to the European Commission. The big pivot away from rule-setting came in a “digital omnibus” proposal on Wednesday — a core part of Commission President Ursula von der Leyen’s “simplification” program to cut red tape to make Europe more competitive. The digital omnibus was one of the “main discussion points” at a meeting between the EU’s tech chief Henna Virkkunen and U.S. Commerce Secretary Howard Lutnick and Trade Representative Jamieson Greer. | Nicolas Tucat/AFP via Getty Images “Whether you call it ‘simplification’ or ‘deregulation,’ you are certainly moving away from the high watermark era of regulation,” said Anu Bradford, a professor at Columbia University who coined the term “Brussels Effect” in 2012. The deregulation drive followed a year in which the Trump administration pressured the EU to roll back enforcement of its tech rulebooks, which Big Tech giants and Trump himself deem “taxes” targeted at U.S. companies. The digital omnibus was one of the “main discussion points” at a meeting between the EU’s tech chief Henna Virkkunen, U.S. Commerce Secretary Howard Lutnick and Trade Representative Jamieson Greer on Monday. “We adopted a major package that would have an impact not only on EU companies, but also on U.S. companies, so this is the appropriate moment … to explain what we’re doing on our side,” European Commission spokesperson Thomas Regnier told reporters on Monday when asked why Virkkunen had discussed the topic with her U.S. counterparts. Lutnick, however, told Bloomberg that Washington was seeking more than just an explanation of EU laws — it wanted changes to its tech rulebooks as well. U.S. giants like Google and Meta have led a full-frontal lobbying push to replace heavy-handed EU enforcement with lighter-touch rules. Behind the push to break the shackles for tech firms is a fear of missing out on the promised economic boom linked to AI technologies. The bloc has traded its role as global tech cop for a ticket to the AI race.  GLOBAL FIRST Brussels showed its ambition to lead the world in regulating the online space throughout the 2010s. In 2016 it adopted the General Data Protection Regulation. Since then, the law has been copied in new legislation across more than 100 countries, said Joe Jones, director of research and insights at the International Association of Privacy Professionals.  When the GDPR came into force, international companies like Microsoft, Google and Facebook acknowledged it spurred them to apply EU privacy standards globally.  It served as a quintessential case of the Brussels Effect: When setting the bar in Brussels, multinational firms would roll out standards across their businesses far beyond the EU’s borders. Other governments, too, copied some of Brussels’ early attempts at setting the rules. After the GDPR, the EU adopted other laws that had the ambition of reining in Big Tech, either by pressing platforms to police for illegal content through its Digital Services Act or by blocking them from using their dominance to favor own services through the Digital Markets Act. Right after the EU adopted its risk-focused AI rulebook, Trump took office and scrapped AI safety rules embraced by his predecessor Joe Biden.  | Chip Somodevilla/Getty Images The EU’s latest blockbuster tech rulebook, the Artificial Intelligence Act, was Brussels’ latest attempt at pioneering legislation, as it sought to address the risks posed by the fledgling technology. “There was more confidence in the EU’s regulation, partially because the EU seemed confident. Right now, when the EU seems to be retreating, any government around is also asking the same question,” Bradford said. Right after the EU adopted its risk-focused AI rulebook, Trump took office and scrapped AI safety rules embraced by his predecessor Joe Biden.   The changing of the guard in Washington came right as Brussels was waking up to the need to be competitive in a global technology race. Former Italian Prime Minister Mario Draghi presented the EU’s competitiveness report in 2024, just weeks before Trump won a second term. “I think the Brussels effect is still alive and well. It just has a bit of the Draghi effect, in that it has a bit of this geopolitical innovation, pro-growth effect in it,” said IAPP’s Jones.  According to German politician Jan Philipp Albrecht, a former European Parliament member who was a chief architect of the GDPR, Europe has become blind to the benefits of its regulatory regime that set the gold standard. “Europeans have no self-secureness anymore … They don’t see the strength in their own market and in their own regulatory and innovative power,” Albrecht said.  WASHINGTON EFFECT Other critics of deregulation are taking a step further, claiming that Washington has hijacked the Brussels Effect — but just on its own terms.   “In an odd way, maybe the Trump administration has taken inspiration from the Brussels Effect, in the sense [that] they see what it means for this one regulating entity to be the one that sets global standards,” said Brian J. Chen, policy director at nonprofit research group Data & Society. It’s just, “they want to be the ones setting those standards,” Chen said. The Trump administration pressured Brussels to tone down its tech regulation during heated trade talks this summer, POLITICO previously reported. That the EU followed through with scaling back its tech laws just as the U.S. is pressing the EU is bad optics, said Schaake, the former lawmaker. “The timing of the whole simplification [package] is very bad,” she said.  She argued that it’s essential to deal with the unnecessary burden on companies, but issuing the digital omnibus after the U.S. pressure “looks like a response to that criticism.” Commission spokesperson Thomas Regnier dismissed the idea that the EU was acting on U.S. pressure. “On the digital omnibus, absolutely no third country had an influence on our sovereign simplification agenda. Because this omnibus is about Europe: less administrative burden, less overlaps, less costs,” Regnier said in a comment on Friday. “We have always been clear: Europe has its sovereign right to legislate,” Regnier added. “Nothing in the omnibus is watering down our digital legislation and we will keep enforcing it, firmly but always fairly.” This article has been updated to include new developments.

BRUSSELS — The EU’s push for the U.S. to scrap its tariffs on steel and aluminum has opened the door to an old demand from Washington: Loosen your digital rulebook, and we’ll meet you halfway. Brussels raised its concerns over Washington’s expanded list of goods covered by high steel and aluminum tariffs at meetings on Monday between Trade Commissioner Maroš Šefčovič and EU trade ministers and, from the U.S. side, Secretary of Commerce Howard Lutnick and Trade Representative Jamieson Greer. The Commerce Department in August subjected over 400 products containing steel and aluminum to a 50 percent tariff — a list the EU feels is so broad it goes against the spirit of a framework trade deal struck in July. That trade deal, which President Donald Trump and European Commission President Ursula von der Leyen clinched at Trump’s Turnberry golf resort in Scotland, sets a baseline tariff of 15 percent on most EU imports to the U.S., while the EU committed to cutting most of its own tariffs to zero. At the time, the EU and the U.S. pledged to work together to reduce tariffs on steel and aluminum — but remained vague on the details. After the Europeans raised the steel tariffs on Monday, Lutnick responded by calling on the EU to “analyze their digital rules, trying to come away with a balance … not put them away, but find a balanced approach that works with us.” “And if they can come up with that balanced approach, which I think they can, then we will, together with them, handle the steel and aluminum issues and bring that on together,” he added. Lutnick’s remarks signal a departure from the previous U.S. position, which threatened to retaliate against the bloc’s digital laws, while advocating for light-touch artificial intelligence regulation.  Lutnick sold the loosening of the bloc’s digital rules as an “opportunity” for the EU, offering U.S. investment in return, mainly through data centers that could power artificial intelligence.   “If the European Union can find a way to have a balanced digital set of rules, I think the European Union can see $1 trillion of investment,” he said. PUSHING BACK — SORT OF In response, Šefčovič reiterated the bloc’s commitment to its regulatory autonomy and its belief that its rules are not — contrary to what Washington asserts — discriminatory.  The EU side, he added, “explained how our legislation is working, we explained that this is not discriminatory. It’s not aimed at American companies. And I think that we just simply need to do more of the explanation in that regard.”  A Commission official, speaking on condition of anonymity, was more direct: “Steel and digital are completely unrelated. Steel has always been part of the discussions with the U.S. and has been formalized in the joint statement. Our sovereign digital legislation is not up for negotiations.” The EU’s digital rules are a major concern for the Donald Trump administration, and U.S. Commerce Secretary Howard Lutnick raised the matter on a visit to Brussels. | Pool photo by Aaron Schwartz/EPA The EU executive has already moved to simplify its tech rules through a digital omnibus presented last week, an effort that the EU’s tech chief, Henna Virkkunen, raised with Lutnick and Greer at an earlier meeting that day. That omnibus brought major changes to the EU’s GDPR data protection regulation, and also proposed to pause the rollout of a key part of the EU’s Artificial Intelligence Act — a controversial move championed by U.S. Big Tech companies and lobby groups.  European lawmakers and civil society groups have expressed concerns in recent weeks that the Commission’s digital simplification push is meant to placate Washington, a claim the Commission has vehemently denied.   Lawmakers are due to discuss the digital simplification package with the Commission on Tuesday. Last week, the Commission also kicked off a process to review all of its tech rulebooks, which could lead to further simplification efforts.  STEEL TALKS  Washington’s earlier decision to widen the list of steel products facing the 50 percent tariff caused uproar in Brussels, with some European lawmakers arguing that the EU should refrain from lowering its own tariffs on steel until the issue is resolved. In a bid to cozy up to the White House, the EU side on Monday pushed the idea that Brussels and Washington should jointly face up to a common enemy — China — rather than dwelling on their differences.  Danish Foreign Minister Lars Løkke Rasmussen said the two sides had addressed “some of the challenges we are facing together,” such as “overcapacity” and “China’s role in the global economy.” Asked about joint work on overcapacity, Lutnick said such issues are “easy for us to work together, and those don’t take up a lot of time when we’re talking, because when everybody just agrees right away, it’s not very difficult.” Behind closed doors, however, the U.S. stressed to its European counterparts that cooperation on China didn’t mean they would simply give the EU a pass on steel and aluminum tariffs. Šefčovič said a team from Brussels would travel to Washington in the coming weeks to address these issues.

LONDON — The robots are coming for us all — even the parliamentary researchers. British politicians — and the industries seeking to influence them — are increasingly embracing artificial intelligence tools in a bid to make their jobs easier.  But the rise of the emerging tech is prompting big questions about the output and job security of young people working in politics — and the vital ladder into the world of Westminster their entry-level gigs provide. “Across the whole of public affairs, you’ll be able to write and communicate better. I think there’s a positive here,” said Peter Heneghan, a former No. 10 deputy digital communications director and now an AI advocate in the public affairs world.  “The negative side of that is there will be a lot of roles that go alongside it,” he added. “It’s inevitable.”  Politicians and the people supporting them are already jumping on AI to help write everything from books, speeches and media briefings to policy proposals and responses to constituency casework.  In public affairs, it’s already proving useful for all manner of run-of-the-mill jobs, including drafting strategies, press releases, communiqués, timelines and media monitoring.  It’s cutting the need to trawl through large documents like Hansard — the official record of the British parliament — or Westminster’s register of all-party parliamentary groups, a frequent source of influence for lobbyists.  Both sources have hundreds of pages added in each routine update — and entry-level staffers can often be found combing them for insight to brief their bosses or clients.  So far, British officialdom is leaning into the trend. The government’s own AI incubator has even created “Parlex,” a research tool leting anybody with a government email address examine a parliamentarian’s stated position on even the most minor issues in little to no time.  Proponents argue these tools will free up people working in politics to do the kind of work AI simply can’t.  But there are frustrations too.   The only sanctioned AI tool for the majority of parliamentary work, as outlined in House of Commons guidance, is Microsoft’s Copilot, which the government has licensed for internal use. | Algi Febri Sugita/SOPA Images/LightRocket via Getty Images The only sanctioned AI tool for the majority of parliamentary work, as outlined in House of Commons guidance, is Microsoft’s Copilot, which the government has licensed for internal use. The use of other chatbots — including ChatGPT, Elon Musk’s Grok, and Claude — is still frequent in parliament, however, amid some grumbling about the official offering.  In June last year, one MP included in a trial emailed the Parliamentary Digital Service — which oversees tech in the Commons — to fume that they “do not want my staff to spend time testing Copilot when the productivity tools are not those that we want or need,” according to correspondence obtained under freedom of information by POLITICO Pro.  ROBOTS TALKING TO ROBOTS  Parliamentarians in the digital age are already inundated with correspondence over email. And artificial intelligence could turn that deluge into an unmanageable flood.  AI-generated email campaigns are now a frequent bugbear for MPs’ offices, with staff feeling pressured to respond to more and more material of a lower and lower quality. One person working in public affairs called it “slop campaigning.”  Heneghan suggests that the “sheer volume” of constituency correspondence that MPs are now getting — and the need to sift through it and reply — means the future of interacting with parliamentarians could become “AI talking to AI.” It would, he says, be “awful” for an already record-low trust in politicians.  Tom Hashemi, the boss of comms consultancy Cast from Clay, echoed that concern. “It’s almost insulting to the point of democracy. MPs are there to respond to genuine constituent concerns, not to have to spend hours of their time responding to AI-generated messages.”  He added that, in his own conversations with ministers and MPs, “they always say those campaigns” — labelled “clicktivism” by Labour MP Mike Reader — “don’t work.”  One parliamentary staffer said: “I can tell that now lots of the email campaigns [by charities] are written by AI — the ones that we get in — whereas before they weren’t. They want it to seem like lots of people are, so they use AI to change the subject lines in the first line of the email very slightly, and the language is all bizarre.”  SQUEEZE ON JOBS AI’s widening use in politics comes amid an increasingly difficult job market for U.K. graduates across the board.  Heneghan suggests there will be a “massive squeeze” on junior jobs available for people working in public affairs, which he argues represents a “double-edged sword” in that menial tasks can be performed more efficiently — while the gains that young people themselves could make from performing them will also be lost.  Prospective job losses will, he predicts, go further than just junior level jobs, with roles for middle managers, human resources, sales and more all being affected.  Meanwhile, Hashemi suggests a route for public affairs firms to continue to expand would be to train new hires to use AI, saying the tech will “affect junior public affairs jobs in firms that don’t adapt to using it and integrating it.”  As trivial as these jobs can seem, many a high-flying politician or adviser got their start shifting around a lot of paper. None other than the prime minister’s chief of staff Morgan McSweeney, for example, got his start in Labour’s headquarters entering data into spin doctor Peter Mandelson’s famed “Excalibur” rebuttal machine.  Current parliamentary aides expressed less concern that AI is coming for them just yet.  Almost all those POLITICO spoke to in parliament said they wouldn’t use AI to write speeches for their bosses, because it is too easy to spot.  However, a Conservative adviser said they imagined junior staffers could become “checkers” of work as opposed to creators of it, due to the ease of asking AI to generate a first pass at materials.  Meanwhile, a second parliamentary staffer said: “It’s like an aid. I don’t think it can replace jobs yet.”  AI’s one attempt to imitate an MP has so far have been widely derided. Labour MP Mark Sewards became the first parliamentarian to create an AI version of himself that constituents could speak to at any hour — to mixed results. It garbled a Guardian reporter’s Northern accent into unintelligibility, and offered relationship advice, alongside producing a deficient haiku about Nigel Farage to PoliticsHome.  That might be the case right now. But as AI continues to develop at breakneck pace, it could soon seem like child’s play. 

LONDON — Britain’s Department of Health is pressing ahead with plans to open up a trove of pandemic-era patient data to outside researchers — despite concerns from doctors’ representatives. A formal direction titled “GP Data for Consented Research,” yet to be signed by Health Secretary Wes Streeting but shared in draft format with doctors’ reps, would enable NHS England to disseminate patient data originally collected solely for the purpose of Covid-19-related research to other studies. The Department of Health and Social Care (DHSC) confirmed to POLITICO that the direction has been drafted and is awaiting Streeting’s signature. A group of doctors has warned the government that the move could erode patient trust. While the direction says government will obtain patient consent to share the data more broadly, doctors groups are worried this won’t happen in practice, and that patients won’t be aware their data is being funneled to other studies. NHS England has been in discussions with the Joint GP IT Committee, which comprises representatives from the British Medical Association (BMA) and Royal College of General Practitioners (RCGP), about the data, a person close to the talks told POLITICO. DHSC confirmed it had been in dialogue with the doctors’ groups, and a spokesperson said it had delayed signing the direction in order to engage with doctors’ concerns. The JGPITC argued it hasn’t been properly consulted on the change in line with established governance processes, and that repurposing the dataset without asking patients’ permission risks damaging already-fragile public confidence in the profession, the same person said.  While the direction says government will obtain patient consent to share the data more broadly, doctors groups are worried this won’t happen in practice, and that patients won’t be aware their data is being funneled to other studies. | Pool photo by Hannah McKay/EPA It comes after the same group of doctors filed a formal complaint to the Information Commissioner’s Office in June alleging that NHS England had breached data protection law by training a general-purpose AI model on the same dataset without consent. The disagreement is also set against the wider backdrop of a long-running dispute between government and the BMA over doctors’ pay and working conditions. DHSC maintains that proper processes have been followed. “As the Secretary of State made clear last year during his speech to the Royal College of GPs in October 2024, we are committed to implementing this direction in line with patients’ explicit consent for their data to be used in research,” a DHSC spokesperson said. ‘CONSULTED EXTENSIVELY’ In his speech last month, Streeting said he would direct NHS England to take responsibility for sharing patient data with projects including UK Biobank, Genomics England and Our Future Health. “I know there are issues we need to work through together around information governance, risk and liabilities,” he said. “There’s also, let’s be honest, some producer interest in play.” NHS England asked the JGPITC to confirm whether it was happy with the direction on broadening access to the dataset by Nov. 4. The JGPITC couldn’t reach a consensus to give its blessing to the change, the same person close to the talks and cited above said. The doctors’ group has pushed for NHS England to notify consenting participants about where their data is going via text or the NHS App, they added. DHSC is not obligated to comply with any of the JGPITC’s requests. “We have consulted extensively with GP representatives over the past 18 months to ensure patients’ wishes are respected and their data used appropriately, while minimizing the burden on busy GPs,” DHSC’s spokesperson said.