Palantir CEO Alex Karp proposed a new regulatory framework for artificial intelligence during an interview on CNBC on September 17, 2026. He advocates for holding individual developers and executives civilly and criminally liable for creating AI systems that cause serious societal harm.
Karp rejects both the European Union's approach to regulation, which he views as stifling to competitiveness, and the libertarian stance of zero oversight. Instead, he suggests applying principles similar to product liability law directly to algorithmic development.
This proposal follows a series of initiatives by Palantir to influence AI governance. In July 2026, the company published a manifesto on AI sovereignty, emphasizing enterprise control over data and infrastructure rather than reliance on third-party platforms.
Karp has also been promoting concepts from his 2025 book, The Technological Republic. In April 2026, he emphasized that the United States must maintain technological supremacy in AI over China, warning that poor regulatory choices could provide strategic advantages to adversaries.
Current industry practices rely on corporate-level governance, such as ethics boards and internal reviews, which largely shield individual employees from legal consequences. Karp argues that personal liability would fundamentally shift risk tolerance for data scientists and engineers.
For investors, this shift could impact the sector significantly. Companies like Palantir, which already maintain robust internal governance, would likely face lower transition costs if individual liability for AI development becomes law.
