An Anthropic investor turns on the AI regulation his peers are pushing for

Joe Lonsdale — an investor in Anthropic and co-founder of Palantir — has publicly pushed back against the intensity of the AI regulation demanded by some of the biggest labs in the field, including some he personally backs. His argument: rules modeled on the safety standards of the dominant players are less likely to protect the public than to entrench the position of those who proposed them, at the expense of competition and innovation.

To an economist, this mechanism has a name: the "raising rivals' costs" strategy, formalized by Steven Salop and David Scheffman as early as 1983. A dominant firm doesn't need to exclude competitors from the market directly — it only has to back a standard, a mandatory audit, or a compliance burden it can absorb effortlessly thanks to its scale, but which falls disproportionately on smaller or newer rivals.

The angle matters directly to the sectors Cardan-AI tracks. In aerospace and defense, where Palantir and Anduril already hold strong positions, an AI compliance standard calibrated to a major player's resources becomes a very real obstacle for a tier-2 or tier-3 subcontractor. The same logic applies to an AI software vendor serving O&G, energy, or luxury: every new requirement deserves one simple question — who is backing it, and who can already meet it painlessly?

The practical takeaway isn't to reject AI regulation on principle — some risks justify a baseline of rules regardless of who benefits. It's to never adopt a standard voluntarily and early without first checking whether its compliance cost is genuinely competitively neutral — or whether it quietly reproduces, under the banner of safety, the structural advantage of whoever pushed for it.