Now that the dust has settled after two extremely eventful days, it's time to take a step back and figure out what this all means for EU digital policy. A few considerations, with the caveat that we probably have more questions than answers for now. Brace for a long thread. 1/18

On Monday, Brussels woke up flat-footed with the resounding resignation of Thierry Breton. Once an all-powerful Commissioner, Breton turned from a rising star to a supernova overnight. His departure leaves a massive power vacuum. My full commentary 👇2/18
content.mlex.com/#/content/1595…

As anyone following EU affairs could not avoid notice, Mario Draghi unveiled his long-awaited report today. He touched upon various issues, but digital technologies in general and AI in particular stand out as a make-it-or-break-it matter. A 🧵on what you need to know. 1/10 The report's focus is on Europe's competitiveness. For Draghi, the origin of the productivity gap between the EU and the US that started to widen in the mid-1990s is explained mainly by Europe's failure to capitalize on the first digital revolution driven by the internet. 2/10

Yesterday evening, an agreement was found on the Council of Europe’s #AI Convention, the first international treaty on AI. The contentious issue of the scope was largely bilaterally negotiated between the European Commission and the U.S. delegation. A 🧵 1/8 The final wording is a solution à la carte. The Convention will cover activities within the lifecycle of AI systems that have the potential to interfere with human rights, democracy & the rule of law when undertaken by public bodies and private actors acting on their behalf. 2/8

I got my hands on an early version of the Commission's recommendation on submarine cables. For the few who, like me, are passionate about the geopolitics of subsea cables, the initiative is particularly interesting as the EU is about to shoot itself in the foot. A 🧵1/12 For background, the security of submarine infrastructure jumped on top of the agenda with the Nord Stream sabotage in 2022. Security is a national competence, but smaller member states are struggling to ensure the security of submarine cables in their territorial waters. 2/12

Now that the trilogue madness is over, here are a few personal reflections on the #AIAct. I should start by saying that I was genuinely overwhelmed by the warm reaction at the press conference. That's a recognition that does not happen every day for a journalist. The AI Act was presented two days after I joined @Euractiv. Thus, it has accompanied me throughout my journalistic career. It's the first landmark law I've followed from start to finish, and, in my view, it is the most important digital legislation of this mandate.

#AIAct trilogue: the second round of the negotiations is starting now, with a showdown expected on law enforcement and national security. Watch this space for some updates in real time. #AIAct trilogue: The EU Parliament has closed ranks against the 🇪🇸package after it was clear it wouldn't have a majority. MEPs developed a counter-proposal, much closer to the parliamentary mandate, with some concessions. The presidency is now analysing. Expect another marathon.

#AIAct trilogue: As EU policymakers prepare for this unexpected second round, a 🧵on the remaining dividing lines and the positions of the actors involved. Spoiler alert: Much will depend on co-rapporteur @brandobenifei and the EU Parliament's centre-left. 1/8
📷@FEPS_Europe There are still around 10 open issues on the agenda, but the real hurdles are the prohibitions, Remote Biometric identification & national security exemption. Yesterday, the 🇪🇸presidency gave MEPs a package deal close to the Council's position, with huge pressure to accept. 2/8

#AI Act trilogue: A small thread on where things landed on foundation models and governance. Take it with a pinch of salt, as the agreement is provisional, and access to sources is still limited as the meeting is ongoing. #AIAct The tiered approach was maintained. The quantitative threshold set on the computing power used to train the model is 10~25. A new annexe to provide criteria for the AI Office to make qualitative designation ex officio or based on a qualified alert from the scientific panel.

Today is D-day for the #AI Act. There is a lot of pressure to reach an agreement, and the stakes are very high. I have little doubt some sort of deal will be found, but there are several pitfalls to watch out for and constraints on all sides. A 🧵 on what you need to know. 1/12 The list of open issues is still mind-boggling, with as many as 23 points. Even if an agreement is reached on all of them, it won't be the end of the story. Intense technical work will follow. Not only to fine-tune the text but also to settle political questions. 2/12

There is nothing better than some Friday drama to close such a hectic week. A technical meeting on the #AI Act has broken down today, but make no mistake: the issue is deeply political and, if no solution is found soon, the whole law is at risk. A 🧵1/8
euractiv.com/section/artifi… At the last trilogue in October, the co-legislators agreed to a Commission's proposal for a tiered approach, which combined transparency obligations for all foundation models and a stricter regime for 'high-impact' ones. This approach was then put black on white. 2/8