Vatican AI adviser warns of cartel behavior among major labs

By Splitfeed EditorialLast updated Sep 26, 2026, 19:17 UTC

On September 23, 2026, Paolo Benanti, a Franciscan friar who serves as the Pope's primary adviser on artificial intelligence, gave an interview to Wired in which he raised two connected concerns about the current state of AI development. First, he argued that major AI laboratories are exhibiting cartel-like behavior, a term that points to concentrated corporate power limiting genuine competition in the sector. Second, he contended that public discourse has become so absorbed by extreme scenarios of godlike AI destroying humanity that more immediate governance questions are being sidelined. Benanti has previously contributed to United Nations discussions on AI governance, giving his remarks an institutional weight that extends beyond the Vatican.

The significance of his comments lies in the distinction he draws between existential risk debates and structural power questions. Discussions about whether advanced AI could pose a civilizational threat have dominated policy conversations for several years, and Benanti's framing suggests that focus may be serving a particular function, intentionally or not, by keeping attention away from how a small number of powerful laboratories are shaping the technology with limited public input. Whether concentrated market behavior in AI genuinely resembles a cartel in any legal or economic sense is a substantive point of dispute, as is the question of whether existential risk concerns are a distraction or a legitimate priority running in parallel with governance work.

None of the three camps have published reactions to this story yet. The Pro-AI camp would typically argue that large labs are competing fiercely and that consolidation reflects the enormous capital requirements of frontier AI, not anti-competitive coordination. The Anti-AI camp would likely find Benanti's cartel framing vindicating, treating it as confirmation that corporate capture of AI development is already well advanced. The Middle Ground camp would probably welcome the call for structured public governance debate while questioning whether the cartel label is precise enough to be useful in policy terms.

The argument Benanti is making will gain or lose traction depending on whether international bodies, including the United Nations forums he has contributed to, move toward binding governance frameworks in the coming months. Any concrete regulatory proposal that references market concentration in AI development would serve as a measure of how far this framing travels beyond a single interview.