Observed Signal · Sep 17, 2026 · Opinion/Editorial · Source: Gary Marcus · Impact: 2/5 · Sentiment: Neutral
AI Liability vs Regulation: Debating the False Dichotomy
Gary Marcus, in an opinion piece, argues against the false dichotomy that AI liability and regulation are mutually exclusive. He criticizes technolibertarians like Joe Lonsdale and David Sacks for using liability arguments to oppose regulation, pointing out that liability alone is insufficient to address AI-related harms. Marcus highlights support from Senator Josh Hawley for both liability and regulation, and cites Mark Cuban's skepticism about passing new laws. He references his 2023 Senate testimony where he argued for a dual approach. The article calls for a comprehensive regulatory framework alongside liability to ensure AI safety, paralleling the aviation industry's model.
Relevant to AI policy but does not introduce new factual events, primarily opinion-driven.
Track OpenAI Signals & Market Shifts in Real-Time
Polaris7 autonomous intelligence agents track regulatory filings, primary sources, executive changes, and deal flow 24/7. Create your free Explorer workspace to monitor these entities.
Key Takeaways & Evidence Grounding
- Gary Marcus argues that AI liability and regulation are not mutually exclusive.
- Senator Josh Hawley supports both liability and regulation for AI companies.
- David Sacks and Joe Lonsdale argue for liability but against regulation.
- Mark Cuban questions the feasibility of passing new AI-specific laws.
- Marcus testified before the U.S. Senate in May 2023 on AI oversight.
- The article references a debate on liability vs. regulation in AI, citing Section 230 ambiguity.
Connected Companies & Entities
1 Entity mapped“Mentioned in the context of security incidents and CEO statements, though not a primary subject....”
Ontology Mapping & Concepts
Related Market Signals & Shifts
Recent verified developments and strategic activity across this market segment.
AI Leaders Face Criticism Over Safety, Regulation Debate
In a Substack post, Gary Marcus criticizes recent public statements by Sam Altman, Jensen Huang, and Bernie Sanders regarding AI safety. Marcus argues that Altman's assurances of OpenAI's trustworthiness are undermined by past statements and the release of GPT-6 Astra, which he claims is less monitorable and safe. He refutes Huang's confidence in self-regulation, pointing to OpenAI's actions. Marcus also disputes Sanders' claim that AI is more dangerous than nuclear weapons, citing casualty estimates. He highlights emerging bipartisan AI regulation efforts by Senators Hawley and Blumenthal, and a cybersecurity proposal by former White House officials Asad Ramzanali and JB Branch, urging focus on practical policies over hyperbolic industry and political claims.
AI Regulation Debate: Congress Blocked, States Advance
This opinion piece critiques the US federal government's inaction on AI regulation, contrasting it with the iterative 'rough consensus and running code' approach that built the internet. It argues that Congress, led by Speaker Mike Johnson, is using industry disagreement as an excuse not to legislate, while the White House, via AI czar David Sacks, champions self-regulation. The article highlights the EU's AI Act, state-level initiatives in 45 US states, and Colorado's recent repeal-and-replace of its AI law as examples of regulatory iteration. It also warns that poorly designed interfaces, as seen with GDPR consent banners, can undermine regulatory intent, and it draws parallels to the social media self-regulation failures exposed by Frances Haugen. The author calls for Congress to act as a competent 'product owner,' shipping narrow, honest rules and revising based on real-world impact.
Section 230 May Not Shield AI Firms from Liability
Gary Marcus argues that Section 230’s immunity for online platforms may not protect AI companies from liability for their models’ outputs, citing a recent German court decision that treated chatbot statements as the company’s own speech. Marcus notes a bipartisan group of U.S. senators has introduced a bill to “sunset” Section 230, and contrasts prior testimony by Sam Altman with OpenAI’s later policy positions and lobbying. He warns that if U.S. courts or lawmakers adopt the German approach—holding vendors responsible for hallucinations, defamation, or harmful advice—major LLM providers (Google, OpenAI, Anthropic, Microsoft, xAI) could face substantial legal risk, which would force engineering and governance changes in AI systems.
Track Real-Time Market Signals & Shifts
Set up custom watchlists to receive automated, evidence-grounded executive digests whenever material signals or shifts occur across your tracked landscape.