US Tightens Grip on Advanced AI, Europe Seeks Digital Sovereignty
The White House has recently imposed significant restrictions on the release of advanced artificial intelligence models by US tech giants like Anthropic and OpenAI, signaling a new and more assertive era of government control over the sector.
What happened
In recent weeks, the US administration halted the rollout of new, cutting-edge AI models, generating uncertainty and debate. Initially, Anthropic was forced to suspend access to its most advanced model, Mythos, only to restore it for a select group of US government organizations and companies after intense negotiations with the White House Wired AI. Almost simultaneously, OpenAI had to delay the launch of its highly anticipated GPT-5.6 models, at the explicit request of the Trump Administration Wired AI.
These moves reflect growing government concern over national security and the potential impact of increasingly powerful AI. In parallel, the legal landscape is becoming more complex: the New York Times has escalated its lawsuit against Microsoft and OpenAI, accusing the former of building a supercomputer to help the latter infringe copyrights. This legal action gains new weight following a recent US Supreme Court ruling against Sony, which has recalibrated the debate on the limits of "fair use" in AI model training Ars Technica AI.
Faced with these restrictions and a growing "nationalization" of AI in the United States, Europe is expressing clear frustration. The continent, already a leader with the EU AI Act for regulation, is now pushing to develop its own sovereign AI models. The goal is to reduce dependence on American and Chinese technologies, ensuring that European values of ethics, privacy, and transparency are intrinsic to internally developed AI systems Wired AI.
Why it matters
The White House's decisions and Europe's reactions are not mere technical or commercial matters; they represent a turning point in global AI governance. Limiting access to advanced models, even if motivated by security concerns, can slow global innovation and create a divide between nations that have access to these technologies and those that do not. This could have profound implications for economic competitiveness, scientific research, and even geopolitical stability.
For businesses and citizens, this means that access to cutting-edge AI tools may not be universal, but conditioned by national policies. The copyright debate, on the other hand, underscores the urgent need for a clear legal framework that balances generative AI innovation with the protection of intellectual property. Without definitive answers, legal uncertainty can stifle investment and confidence in the sector.
The HDAI perspective
These events reinforce the conviction that technology always serves human and political decisions. The race for AI supremacy, coupled with the need for regulation, highlights the urgency of an approach that places humanity at its center. It's not just about building more powerful models, but about ensuring these models are developed and used ethically, transparently, and responsibly.
For Human Driven AI, the issue of European technological sovereignty is crucial. A strong and independent Italy AI summit and European AI can act as a counterweight to AI dominated by a few superpowers, promoting a more diverse ecosystem based on shared values. This is a central theme we will address at the HDAI Summit 2026, where experts and leaders will discuss how to balance innovation, security, and genuine ethical AI for the common good. AI governance is not a technical problem, but a challenge of values and global cooperation.
What to watch
It will be crucial to observe how AI control policies evolve in the United States and what Europe's next steps will be to realize its ambition of developing sovereign AI models. Future rulings on copyright will also define the legal boundaries for training and using generative models, directly influencing the future of digital innovation and creativity.