Speaking at the Reuters Momentum AI Austin event, Joe Lonsdale, co-founder of Palantir and founder of 8VC, accused Anthropic and OpenAI of using concerns over existential AI risks to influence public policy.
He warned that regulations championed by leading AI labs could make it harder for smaller competitors to keep pace.
Lonsdale said: “I think that what OpenAI and Anthropic are pushing right now is very dangerous, and we don’t want an oligopoly that controls all of our policy here with the government.”
The criticism comes as the AI industry’s biggest names publicly called for self-regulation. Anthropic CEO Dario Amodei, OpenAI CEO Sam Altman and Elon Musk argued in early September for tighter controls on advanced AI systems, warning that capabilities are advancing faster than existing safeguards.
Lonsdale is not the only tech leader to push back against the call. David Sacks, the Trump administration’s AI and crypto czar, has accused Anthropic and OpenAI of establishing a “duopoly on frontier intelligence” and of trying to push out smaller, less established developers by urging for regulations.
“Safety is a real thing, but fear and regulation – we’ve got to be really careful how we respond to those things,” Lonsdale told Reuters.
The push for caution is taking place against a backdrop of ever-accelerating AI development. Anthropic launched Claude Opus 5.5 on 22 September – a week after calling for brakes on advancement – while leading labs continue investing billions in infrastructure and model development. Meanwhile, the public and governments are also calling for restraint.
Safety concerns are driven by reports of increasingly autonomous AI behaviour. AI agents have successfully hacked external websites by escaping testing environments and interacting with public systems in unexpected ways.
As more incidents arise of advanced AI systems accessing private business and government data without developers’ knowledge, vocalised concerns about cybersecurity and oversight are becoming louder.
Policymakers face growing pressure to establish testing and governance frameworks for frontier models, particularly those capable of autonomous decision-making or destructive cyber activity.
The debate is also shaped by geopolitics. While public support for AI regulation is growing, the US and China are simultaneously racing for leadership in the technology. When pushing back against his country’s AI pioneers, president Trump said, “We’re leading China in AI… and frankly, I want to keep it that way, because whoever wins AI wins.”
China’s response to the US companies’ call for slowed development was resoundingly negative; some of Lonsdale’s criticism could be heard in their reaction. A spokesperson for Beijing’s Foreign Ministry accused the tech giants of sharing “threat narratives and engaging in confrontation and malicious competition.”
Despite this, Trump hosted Chinese president Xi Jinping alongside several technology leaders at a state dinner last week to discuss the AI race. The meeting appears to have been friendly, with Xi announcing both countries were responsible for managing “AI for good” and for ensuring that “AI is always under human control and serves the well-being of the people.”
As AI development inevitably continues, regulators must balance safety concerns with fears that restrictions could slow domestic innovation and strengthen foreign competitors. Lonsdale’s intervention highlights the central tension facing the sector, especially as the criticism has come from an Anthropic investor directly.
Governments are under pressure to introduce safeguards as AI becomes more capable. However, the question is whether those safeguards will improve safety across the industry or reinforce the dominance of the companies already leading the AI race.
RELATED STORIES
Capacity Europe 2026
The 25th anniversary edition of Capacity Europe 2025 will bring together 3,500+ decision-makers from the global connectivity and digital infrastructure community.
