Arthur Mensch, CEO of Mistral AI, cautioned at the VivaTech conference in Paris that the most serious risk of artificial intelligence may not be mass job loss—but rather deskilling, as humans become overly dependent and mentally complacent.

Mensch, whose startup has quickly grown into a major European AI presence, said that while some experts warn AI could displace white-collar workers, he believes the real threat is people relying too heavily on AI and failing to think critically.

“The biggest risk with AI is not that it will outsmart us or become uncontrollable, but that it will make us too comfortable, too dependent, and ultimately too lazy to think or act for ourselves,” Mensch stated.

He urged AI developers to include active human participation in model design—ensuring users review and challenge AI output rather than passively accepting it as truth.

Mensch also downplayed fears of widespread white‑collar unemployment, saying job roles will shift—not vanish—and noted that relational tasks requiring empathy and human judgment will become more valuable. His comments push the industry to rethink how automation is implemented and remind us that learning and critical thinking must stay central in the age of AI.