Can AI Outpace Governments? Luis Bravo Debates the Future of Regulation on The Officer
Our Head of Information Security —who represented Spain at the UN headquarters in New York this year— joins The Officer podcast to discuss the US veto on some of Anthropic's AI models, technological sovereignty, and whether Europe is ready for this race. Watch the full episode here.
Madrid. Artificial intelligence is no longer just a technological matter — it's a matter of power. Do the most advanced AI models pose a real threat to national security? Should companies depend on a single AI provider? Is Europe ready to compete in this new race? These were the questions debated by Luis Bravo, our Head of Information Security, on The Officer podcast, alongside Andrés Mendoza, Technical Director for Southern Europe.
You can watch the full episode on YouTube.
The trigger: the US veto on Anthropic's models
The conversation starts from an event that has shaken the industry: the recent US veto on some of Anthropic's AI models, which has reopened the debate on regulation, national security and technological sovereignty. It's a subject we know well — we already analysed the impact of Claude Mythos and the vulnerabilities in AI-generated code when models with offensive capabilities proved that cybersecurity had changed in scale. When AI can find vulnerabilities faster than governments can legislate, the question in the title is no longer rhetorical.
Don't marry a single provider
Another key theme of the episode is technological dependence. What happens when your entire operation rests on the model of a single company, subject to the commercial or regulatory decisions of another country? It's a conviction we apply to everything we build: our solutions, such as ADA, are model-agnostic and use the best-performing AI for each task at any given moment — precisely so that our clients are never tied to one provider or exposed to a veto.
Europe, between regulation and the race
The episode also asks whether Europe is ready to compete. Our position is nuanced: the European framework —the AI Act, GDPR, data sovereignty— is not just a burden, but an opportunity to stand out for those who build technology that complies by design. That's the bet we made with ADA, deployable on-premise and designed for regulated environments, and the vision Luis also took to the UN Open Source Week in New York, where he was part of the team that represented Spain after winning the Spanish Government's Responsible AI hackathon.
Hit play
Cybersecurity, regulation, the geopolitics of AI and Europe's role — half an hour of straight talk. Watch the full episode here. And if your company is asking itself any of these questions, let's talk.
Key takeaways
Because of their offensive potential in cybersecurity: models capable of detecting vulnerabilities and generating exploits reopened the debate on national security and access control to the most advanced AI.
Yes. Beyond the technical risk, there's a regulatory and geopolitical one: a policy change or a veto can leave your operation exposed. The alternative is a model-agnostic architecture.
Europe regulates more, but that framework (AI Act, GDPR) can become a competitive advantage for technology that complies by design, as Logixs argues in this episode.