Diritto intelligente – Issue N. 16

Artificial intelligence is no longer living in a legal grey zone. The contributions in this January issue of Diritto Intelligente show a clear shift: AI has moved from experimentation to accountability.

The CNIL’s January 2026 Recommendations on AI and GDPR compliance set the tone. Rather than creating new legal categories, the French authority applies existing data protection principles with renewed rigour. The focus on AI models—not just systems—makes one point clear: anonymity cannot be assumed, and risk cannot be outsourced downstream. Compliance becomes a continuous exercise, not a one-off assessment.

A similar logic runs through the European Commission’s investigation into Google’s use of publishers’ and creators’ content to power generative AI services. AI is no longer treated as a neutral layer of innovation, but as a tool that can reshape visibility, traffic, and revenue in digital markets. Control over data and distribution matters, and competition law is beginning to reflect that reality.

The uncertainty surrounding AI training and intellectual property emerges sharply in Getty Images v Stability AI. The decision to allow an appeal on whether an AI model itself can be an “infringing copy” shows how unresolved the legal boundaries remain—and how high the stakes are for the generative AI ecosystem.

The issue then looks beyond disputes and enforcement, showing how AI is transforming drug discovery through the Recursion case. Here, efficiency gains are real, but they do not eliminate the need for governance, explainability, and regulatory discipline.

Finally, the focus on legal engineering brings the discussion back to organizations. AI tools are everywhere, but value is not automatic. Without the skills to translate legal needs into workable systems, technology risks falling short.

Together, these contributions tell the same story: AI has grown up. And with maturity comes responsibility.