Issue #16 - February 4, 2026

Buy European: industry and Commission show signs of unity

More than a 1,000 European companies have joined an op-ed signed by Stéphane Séjourné to call for a Buy European intiative. This initiative comes a few weeks after the Greenland crisis, and illustrates how industries across the EU have become nervous about relying on foreign technologies and imports without alternatives within the EU market. “Buy European” is part of a broader file led by the French commissioner, called the “Industrial Accelerator Act”. Part of it is a “Made in Europe” plan, that would lay out minimum thresholds for locally produced parts and supplies, but has been criticised by European carmakers, on the basis that it could ultimately increase the price of cars, in a market already struggling with cheap Chinese EV cars flooding the European market. Despite the blowback from the auto industry, Stéphane Séjourné comes with the support of 1,141 business leaders, ranging from pharmaceutical, technology, steel, but also auto parts, to press ahead with his file. The Act would in addition to include minimum thresholds, reward companies with products using high levels of parts and supplies made in the region with public subsidies. While the US has been bullish in its Made in America plan, Europe is tempted to replicate a similar approach with a subsidy-based reward model related to a percentage of materials being made in Europe. While it politically makes sense to apply a mechanism of reciprocity with US rules, many other aspects should be taken into account. First, how to keep up with the heavy innovation race that is happening at global level? While Tesla is moving away from its automobile core focus to invest in physical AI and robotaxis, and Chinese carmakers are already a serious contender of Tesla, which policies will position the European auto industry to stay relevant in the AI and EV race? That might start with anticipating the key technological trends and avoiding market failures for robotaxis, self-driving cars while promoting the emergence of European players. Secondly, re-industrialisation of Europe in a world that moves toward deglobalisation will be costly. Manufacturing dependencies with Asia are too heavy to be downplayed, and strategic alliances such as Mercosur likely will help lower the bill of complex chains of logistics. But first, this will require strategic assessment of dependencies the EU can and cannot afford. Finally, a fully integrated single market for Europe would alone help tremendously. A EU capital markets would enable innovators make the heavy investments needed, a unified energy one, lower the bill for carmakers, etc. but while all of this seems still very far from materialising, EU Inc, an initiative led by startups across the EU and now supported by Von der Leyen is certainly the low hanging fruit that needs more support. The initiative, referred to as “28th regime” in Letta’s report, calls for harmonising legal rules across the EU, and could boost productivity and capacity for startups to thrive in the EU without having to move to the US to succeed.

Youth social media bans gain momentum while AI chatbots reveal new safety gaps

While many European countries have engaged in discussions over how to limit the negative impact of social media on youth, the ban trend ignited by Australia in December has gained much support in Europe. Now Spain has joined France in a call to ban social media for youth. This represents a drastic move that involves complex enforcement of rules with privacy enhanced age verification at its core. The embrace of outright bans over graduated restrictions (age-appropriate content, enhanced parental controls, usage limits) signals governments have concluded voluntary platform cooperation won't deliver adequate protection. While youth policy on social media is still in the making, AI chatbots may very well become a top of mind issue very soon too. An AI child safety benchmark has been conducted by Kora on dozens of AI chatbots, and the results suggest they pose distinct child safety challenges that may soon face similar regulatory scrutiny. While the benchmark shows a wide gap between best in class and worse in class chatbots for safety, education is lagging behind for almost all of them. The benchmark notes that “models fail most often on educational integrity risks [...], particularly academic dishonesty where 76% of responses were inadequate. AI systems are poorly calibrated to protect children from cheating and misinformation in learning contexts”. While educators already have a poor opinion of chatbots, the industry will need to tremendously invest in this area to build confidence among parents, teachers and governments. Given the reputational stakes involved, investing in this area should be a tier 1 priority.