Maximilian Henning @maximilian-henning.eurosky.social · 4hRead my article at @euractiv.com: www.rapporteur.com/news/exclusi...rapporteur.comEXCLUSIVE: China offered European Commission to join its AI alliance | RapporteurEU executive is already part of rival US initiative to secure AI development 000 0
Maximilian Henning @maximilian-henning.eurosky.social · 4hBoth Beijing and Washington are building their own AI political blocs. The Commission – on behalf of the entire EU – has signed up to the United States’ Pax Silica initiative, and Washington is warning that no one can be part of both. 100 1
Maximilian Henning @maximilian-henning.eurosky.social · 4hThe Commission informed national experts about the offer this week. No word yet on when exactly China made it or whether the EU executive has replied. 100 1
Maximilian Henning @maximilian-henning.eurosky.social · 4hChina has offered the European Commission to join its international AI alliance, the World Artificial Intelligence Cooperation Organisation, which it set up earlier this year. 100 1
Maximilian Henning @maximilian-henning.eurosky.social · 19/10/2026Read my article at @euractiv.com: www.rapporteur.com/news/eu-inc-...rapporteur.comEU Inc heads for potential Council deal | RapporteurCountries could give thumbs up to draft position on Friday 010 1
Maximilian Henning @maximilian-henning.eurosky.social · 19/10/2026- Sets out a way for countries with “co-determination” to make sure EU Inc couldn’t be used to subvert workers’ rights - Further limits how long EU Inc companies would be able to use a proposed harmonisation of when stock options get taxed - Adjusts rules for preventive controls 111 1
Maximilian Henning @maximilian-henning.eurosky.social · 19/10/2026How about some good news for #EUInc? Council could agree its position on the project this Friday, I’ve been told. I also got my hands on the new – and potentially final – draft text for that position. It tries to tackle three areas in which capitals had so far still differed: 100 1
Maximilian Henning @maximilian-henning.eurosky.social · 15/10/2026Read my article at @euractiv.com: www.rapporteur.com/news/exclusi...rapporteur.comEXCLUSIVE: Hungary pitches putting EU safety teams inside frontier AI labs | RapporteurPaper also sketches new 'legislative track' to tackle internal model risks and liability 000 0
Maximilian Henning @maximilian-henning.eurosky.social · 15/10/2026Key parts: The EU should place evaluators directly inside frontier AI labs – and consider strengthening its AI Act to explicitly cover top models those labs are running internally. 100 1
Maximilian Henning @maximilian-henning.eurosky.social · 15/10/2026Hungary has put forward a concrete proposal for how the EU could strengthen its AI rulebook, reacting to months of never-ending reports about safety incidents. 100 1
Maximilian Henning @maximilian-henning.eurosky.social · 15/10/2026Another setback for #EUInc: Lawmakers failed to close a deal on the project for a second time on Wednesday night. Dealing with specific rights some countries grant their workers again proved the stumbling block. Read my article at @euractiv.com: www.rapporteur.com/news/eu-inc-...rapporteur.comEU Inc again stumbles over workers’ rights in Parliament | RapporteurLawmakers fail to agree deal on crucial economic project for second time 022 2
Maximilian Henning @maximilian-henning.eurosky.social · 14/10/2026Read my article at @euractiv.com: www.rapporteur.com/news/mistral...rapporteur.comMistral in the running to back one of EU’s AI gigafactories | RapporteurBut Europe can do more with less, says France's AI hopeful 000 0
Maximilian Henning @maximilian-henning.eurosky.social · 14/10/2026Europe needs to compete on building frontier models and can do so with more compute, she argued. But she also criticised that the project was still partly being planned along national lines, which she suggested contributed to its slow progress. 100 1
Maximilian Henning @maximilian-henning.eurosky.social · 14/10/2026Mistral is preparing to take part in an EU push to build AI datacentres – labelled “AI gigafactories” – a representative of Europe’s AI hopeful said in Brussels on Tuesday. 100 1
Maximilian Henning @maximilian-henning.eurosky.social · 13/10/2026In many of the AI security incidents that have recently made headlines, the involved models were lab-internal. A group of advisors has now called on the @ec.europa.eu to make sure there are rules for such cases. Read my article at @euractiv.com: www.rapporteur.com/news/eu-advi...rapporteur.comEU advised to tighten AI law's grip on internal models | RapporteurScientific advisors and MEPs urge Commission to address risks posed by non-public AIs 024 4
Maximilian Henning @maximilian-henning.eurosky.social · 12/10/2026The research is here: bias.latticeflow.aibias.latticeflow.aiLatticeFlow AI Political Bias EvaluationLatticeFlow AI first framework to measure political bias across leading AI models 000 0
Maximilian Henning @maximilian-henning.eurosky.social · 12/10/2026AI models have their own biases for how they answer questions. Elon Musk built Grok to be less “woke”, while Chinese ones toe the communist party line on sensitive topics. But where do European models fit into this? Read my article at @euractiv.com: www.rapporteur.com/news/europea...rapporteur.comEuropean AIs bit more ‘neutral’ than US or Chinese ones, study finds | RapporteurResearch into political bias also shows 'debiasing' Chinese models works quite well 110 1
Maximilian Henning @maximilian-henning.eurosky.social · 09/10/2026Read my article at @euractiv.com: www.rapporteur.com/news/hungary...rapporteur.comHungary claims Orbán-era AI surveillance in line with EU rules | RapporteurCommission keeps mum on whether it shares Budapest's opinion 010 1
Maximilian Henning @maximilian-henning.eurosky.social · 09/10/2026The country might have switched government since then, but its interior ministry is still telling me there is no issue with these EU rules. The @ec.europa.eu did not want to speak about “confidential bilateral discussions” on the topic. 110 1
Maximilian Henning @maximilian-henning.eurosky.social · 09/10/2026Remember when Hungary last year allowed police to use facial recognition on Pride protestors in Budapest? Civil society then argued this wasn’t in line with the EU’s AI Act, which puts strict limits on the tech. 120 2
Maximilian Henning @maximilian-henning.eurosky.social · 06/10/2026A few weeks’ delay might normally not mean much, but in this case Parliament has signed up for wrapping up talks until the end of the year – as did national capitals, which are having their own problems. Read my article at @euractiv.com: www.rapporteur.com/news/eu-inc-...rapporteur.comEU Inc faces setback in Parliament as year-end deadline looms | RapporteurCountries also still lack a common position on the startup-booster reform 011 1
Maximilian Henning @maximilian-henning.eurosky.social · 06/10/2026The #EUInc project is facing a delay in Parliament as lawmakers did not reach an overall deal on Monday night. They do however seem to have gotten pretty close apart from one key issue. 100 1
Maximilian Henning @maximilian-henning.eurosky.social · 05/10/2026Safety incidents at OpenAI have repeatedly made headlines in the past few weeks. I wrote up how the @ec.europa.eu could turn up the heat to pressure the company into improving its safety practices – if it wants to. Read my article at @euractiv.com: www.rapporteur.com/news/how-the...rapporteur.comHow the EU could react to OpenAI’s safety incidents | RapporteurAs AI Act enforcer, Commission could probe models or order changes 001 1
Maximilian Henning @maximilian-henning.eurosky.social · 02/10/2026Some people might think there are no AI model builders outside of the United States and China, but that’s not true. For example, there is Cohere – a Canadian-born company betting strong on Europe in these sovereign times. Read my article at @euractiv.com: www.rapporteur.com/news/canadia...rapporteur.comCanadian-German AI lab eyes building sovereign AI for Europe | RapporteurCohere stands on both sides of the Atlantic after merger with Aleph Alpha 041 4
Maximilian Henning @maximilian-henning.eurosky.social · 01/10/2026I’ve got my hands on a draft law on standardisation, which the European Commission is set to propose next week. It aims at improving Europe’s influence on important technologies, and to speed up the writing of standards. Read my article at @euractiv.com: www.rapporteur.com/news/draft-s...rapporteur.comDraft standards law to boost EU sway over strategic tech | RapporteurCommission also wants to speed up writing of technical norms, per draft law 000 0
Reposted by Maximilian HenningThéophane Hartmann @theophane.euractiv.com · 30/09/2026EXCLUSIVE: France proposes expanding cloud sovereignty law to entire IT sector Expanding the rules would create a massive inflow of public spending into European alternatives @euractiv.com www.rapporteur.com/news/exclusi...rapporteur.comEXCLUSIVE: France proposes expanding cloud sovereignty law to entire IT sector | RapporteurExpanding the rules would create a massive inflow of public spending into European alternatives 002 2
Maximilian Henning @maximilian-henning.eurosky.social · 29/09/2026One part of that will be a “structured, living register” collecting barriers for AI uptake in manufacturing and robotics, set to be launched next month already. Read my article at Rapporteur: www.rapporteur.com/news/exclusi...rapporteur.comEXCLUSIVE: EU to create register of barriers to industrial AI | RapporteurReport on manufacturing and robotics runs parallel to work by European Commission's AI envoy 000 0
Maximilian Henning @maximilian-henning.eurosky.social · 29/09/2026But a document I’ve seen says another Commission department is currently helping national ministries and trade groups out on writing another report with recommendations – basically on the same topic. 110 1
Maximilian Henning @maximilian-henning.eurosky.social · 29/09/2026Ursula von der Leyen has charged a special envoy with finding out how to push Europe’s industrial companies into adopting AI, assisted by the Commission’s in-house thinktank. 110 1
Maximilian Henning @maximilian-henning.eurosky.social · 28/09/2026A coalition of trade unions and civil society organisations wants European governments to end all existing contracts with Palantir. Many countries are already reconsidering their ties to the American data analytics company. Read my article at @euractiv.com: www.euractiv.com/news/kick-ou...euractiv.comKick out Palantir, trade unions tell European governments | EuractivCall to eject US data analytics giant on democratic grounds 082 8
Maximilian Henning @maximilian-henning.eurosky.social · 24/09/2026Trade unions from across Europe on Thursday demonstrated against #EUInc in Brussels. Protestors said they feared social dumping and threats to co-determination. Read my article at @euractiv.com: www.euractiv.com/news/trade-u...euractiv.comTrade unions protest in Brussels against EU Inc | EuractivParticipants fear social dumping and threats to co-determination 064 6
Maximilian Henning @maximilian-henning.eurosky.social · 24/09/2026Thanks to @frederike.bsky.social, @maxvonthun.bsky.social and @leevisaari.eurosky.social for their input! 010 1
Maximilian Henning @maximilian-henning.eurosky.social · 24/09/2026But what if we don’t? What if we end up in a world where models become interchangeable – “commoditised”? What should Europe’s course on AI be then? Read my article at @euractiv.com: www.euractiv.com/news/what-is...euractiv.comWhat is Europe's plan if AI doesn’t self-improve? | EuractivCivil society AI experts are eyeing a ‘commodified’ alternative future 101 1
Maximilian Henning @maximilian-henning.eurosky.social · 24/09/2026A lot of fears and hopes about AI hinge on whether we’ll get “recursive self-improvement” – models building their own successors, a potentially enormous security risk and also a great advantage for leading AI labs. Anthropic’s Dario Amodei said he’s seeing first signs of it. 110 1
Maximilian Henning @maximilian-henning.eurosky.social · 18/09/2026The EU’s AI Office is apparently also talking to leading AI labs, including OpenAI, about how they’re planning to prevent cases like this one from happening again. No verdict yet on whether planned changes are “appropriate”. Read my article at @euractiv.com: www.euractiv.com/news/exclusi...euractiv.comEXCLUSIVE: OpenAI didn’t report safety incident under EU AI rules | EuractivEuropean Commission says it's assessing AI labs’ recent safety improvements 032 3
Maximilian Henning @maximilian-henning.eurosky.social · 18/09/2026OpenAI did not report a recently uncovered safety incident under EU safety rules, the European Commission told me. If you lost track: This is the “RubyGems” incident, first reported last Friday. 140 4
Maximilian Henning @maximilian-henning.eurosky.social · 17/09/2026The world is debating whether AI will kill us all after warnings coming out of American labs. But European and Canadian AI companies are very sceptical – especially of the pitched response. Read my article at @euractiv.com: www.euractiv.com/news/europea...euractiv.comEuropean AI labs hit back at American ‘doomer bubble’ | EuractivProposal for AI giants to coordinate on rules also branded a “cartel” by critics 020 2
Maximilian Henning @maximilian-henning.eurosky.social · 16/09/2026Commission President @vonderleyen.ec.europa.eu has just said she’ll invite leading AI labs for talks on pacing the frontier of the tech. The move comes after days of intense, worldwide discussions around AI safety. Read my article at @euractiv.com: www.euractiv.com/news/eu-chie...euractiv.comEU chief to host safety talks with leading AI labs, per draft speech | EuractivPlanned invite follows global discussion of AI labs' doomsday predictions 020 2
Maximilian Henning @maximilian-henning.eurosky.social · 15/09/2026You want to know about von der Leyen’s State of the Union speech? We have a draft. Read all about it in @euractiv.com’s liveblog: www.euractiv.com/news/live-vo...euractiv.comLIVE: Von der Leyen's SOTEU speech leaked | EuractivEuractiv has seen a draft of Ursula von der Leyen's State of the EU' speech 010 1
Maximilian Henning @maximilian-henning.eurosky.social · 15/09/2026A group of luminaries including Margrethe Vestager is pitching key EU countries on setting up a new alliance pooling their leverage on the world’s AI supply chain – and say it should be linked to the EU’s trade bazooka. Read my article at @euractiv.com: www.euractiv.com/news/europe-...euractiv.comEurope must tighten grip on AI supply chain, experts warn | EuractivVestager among backers of pitch to target EU's trade bazooka at frontier AI access 010 1
Maximilian Henning @maximilian-henning.eurosky.social · 14/09/2026www.euractiv.com/opinion/edit...euractiv.comEditors' Choice: AI Act has come of age and not a moment too soon | EuractivThe AI Act was much maligned for "stifling innovation". Given recent incidents, those complaints look ill-judged 011 1
Maximilian Henning @maximilian-henning.eurosky.social · 14/09/2026I’d be fine with talking less about the rules that AI bosses might want and more those which they already have to follow right now. Read my opinion piece at @euractiv.com: 120 2
Maximilian Henning @maximilian-henning.eurosky.social · 14/09/2026We have rules for AI already. If AI labs think their models are not safe, they shouldn’t proceed with developing them; if they detect serious safety incidents, they have to report them. 110 1
Maximilian Henning @maximilian-henning.eurosky.social · 14/09/2026I wrote this before the weekend, but my thoughts still apply after the bosses of several American AI labs aired their safety concerns over the weekend. 110 1
Maximilian Henning @maximilian-henning.eurosky.social · 14/09/2026There are some questions that AI models built by frontier labs won’t answer – for safety reasons. But getting rid of these built-in safeguards has turned into its own field of business. Read my article at @euractiv.com: www.euractiv.com/news/ai-jail...euractiv.comAI jailbreakers test EU safety rules | EuractivChatbots advertised as having no safeguards are readily available online 010 1
Maximilian Henning @maximilian-henning.eurosky.social · 10/09/2026The @ombudsman.europa.eu has opened an inquiry into the @ec.europa.eu’s appointing of a “special envoy for industrial AI”. The case is specifically about the EU executive not replying to a letter with questions from civil society. Read my article at @euractiv.com: www.euractiv.com/news/eu-watc...euractiv.comEU watchdog probes Commission's AI envoy appointment | EuractivEU's executive asked to reply to questions about Siemens chairman’s role 000 0
Reposted by Maximilian HenningThéophane Hartmann @theophane.euractiv.com · 09/09/2026As @ec.europa.eu publishes new public procurement rules today, @maximilian-henning.eurosky.social, @steporciello.bsky.social, Thomas Moller-Nielsen and I looked at the range of measures designed to favour EU alternatives over US and Chinese suppliers @euractiv.com www.euractiv.com/news/eu-proc...euractiv.comEU procurement blueprints reignite 'European preference' debate | EuractivBrussels weighs quotas, local content rules and other ways to favour EU products 011 1
Maximilian Henning @maximilian-henning.eurosky.social · 09/09/2026This is not the European Commission. The Commission proposed building a register. It's the national capitals which want to get rid of it in Council. 110 1
Maximilian Henning @maximilian-henning.eurosky.social · 09/09/2026Capitals have booted a planned central register for #EUInc companies out of their latest legal draft for the project. Lawmakers have so far kept the register in their take, though this is all of course preliminary. Read my article at @euractiv.com: www.euractiv.com/news/capital...euractiv.comCapitals want to ditch EU Inc's central register | EuractivPlan for a register inside bloc's startup reform law faces fresh pushback 013 3
Maximilian Henning @maximilian-henning.eurosky.social · 08/09/2026OpenAI has NOT submitted a report specifically on the “wiki incident” uncovered by Reuters on Friday to the EU, I have now learned. Model developers have to notify serious AI safety incidents under EU rules for the tech. Read my article at @euractiv.com: www.euractiv.com/news/ai-safe...euractiv.comAI safety incident reporting ‘not just a tick box’, warns EU | EuractivCommission remarks follow OpenAI confirming fresh incident with its AI agents 022 2