Anthropic researcher resigns: "AI is moving faster than our control"

A researcher at Anthropic, a leading company in artificial intelligence development with a mission focused on safety, has resigned by issuing a stark warning. They left their position explicitly stating that AI is advancing faster than its creators' ability to maintain control, describing the risk as "profound" and concerning all of humanity.

The decision, described as "costly" and not symbolic, comes from within one of the most prominent laboratories on the issue of AI model safety. This brings the question of responsibility in the accelerated development of artificial intelligence back to the center of public debate—a topic often discussed but rarely brought to light by a direct protagonist who decides to abandon their work.

Anthropic is an artificial intelligence research company founded with the explicit goal of developing "safe, reliable, and interpretable" AI systems. Its creation was motivated precisely by the concern that advanced AI could escape human control if not designed with adequate safety constraints. This distinguished it as a "cautious" company in a sector known for a frenzied race to develop.

In recent years, competition in the generative AI sector (such as language models) has become increasingly intense, with major tech companies and startups releasing ever more advanced capabilities. This has created commercial pressure that often sidelines long-term safety assessments. The researcher's decision to resign signals that even within a company born to be a bastion of caution, the competitive drive is winning, making the problem systemic for the entire sector.