The Tool Desk
Outbyte Driver Updater FREEFix the driver behind crashes, sound loss and screen glitchesFind Drivers →Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →AI liability and AI regulation serve different purposes: lawsuits can assign responsibility and seek remedies after harm, while regulation can set duties before or during deployment. Gary Marcus argues that accepting liability is no reason to reject regulation—and that litigation alone may be too slow or uncertain to address some AI-related harms.
What Marcus means by the “false dichotomy”
In a commentary essay dated 17 September 2026, Gary Marcus challenges the argument that if AI companies can be held liable, additional regulation is unnecessary. His point is not that liability should be replaced. It is that liability and regulation answer different accountability needs and can coexist.
As an Amazon Associate I earn from qualifying purchases.
The essay states: “We absolutely need to hold AI companies liable for the damage they cause.” Marcus’s accompanying argument is that relying on lawsuits alone assumes existing law clearly covers the relevant harms and that court cases can deliver timely remedies. He disputes those assumptions for some situations. This is Marcus’s policy argument, not an established finding that one particular combination of rules and lawsuits is best.
What lawsuits can do—and where Marcus sees limits
Liability works through claims about particular conduct and harm. Litigation can provide a route to compensation or other legal remedies, and the possibility of being sued can influence decisions. But a lawsuit typically addresses a dispute after events have occurred. Its reach depends on applicable law, the facts, and a claimant’s ability to bring a case.
#1 Best Overall
Marcus’s essay recounts his 2023 Senate testimony in an exchange with Senator Josh Hawley. The essay reproduces Hawley asking, “Why don’t we just let people sue you?” and Marcus responding that lawsuits can take a long time. It also attributes to Marcus the statement, “The laws that we have today were designed long before we had artificial intelligence.” These are quotations as reproduced in the essay, not independently authenticated here against an official hearing transcript.
Marcus points to copyright, mass misinformation, and uncertainty about Section 230 as examples of questions he believes existing legal frameworks may not resolve clearly or quickly. Those examples describe his argument; they are not conclusions here about what current U.S. law covers. The essay does not establish the full status of federal or state AI liability laws, current cases, or agency rules.
What regulation adds
Regulation can establish obligations before or while systems are developed, supplied, or used. Depending on the law, those obligations may include prohibitions, transparency requirements, governance duties, or controls for defined categories of systems. That makes regulation different in timing and form from case-by-case litigation; it does not by itself guarantee effective oversight or eliminate disputes about responsibility.
- Timing: Liability generally addresses a claim after an alleged harm; regulation can set expectations in advance or during deployment.
- Specificity: Courts assess particular claims under applicable legal standards; regulations can define duties for categories of actors or systems.
- Coverage: Both approaches depend on scope. A lawsuit depends on applicable law and the facts; a regulatory duty depends on whether the actor and system fall within the rules.
- Enforcement: Litigation relies on access to courts and the ability to pursue a claim. Regulation relies on regulators and their capacity to interpret and enforce requirements.
These are distinctions, not proof that regulation is always faster, clearer, or more effective. Institutional capacity, oversight, and the risk of regulatory capture matter, and Marcus’s essay does not resolve those trade-offs.
The EU AI Act shows how regulation can be staged
The European Union’s AI Act provides a concrete example of duties taking effect on different schedules. The European Commission describes the framework as phased; these dates apply to the EU and should not be read as a worldwide timetable.
| Date | Milestone |
|---|---|
| 2 February 2025 | Prohibitions and AI literacy obligations began applying, according to the European Commission’s AI Act regulatory framework overview. |
| 2 August 2025 | Governance and general-purpose AI obligations became applicable, according to the European Commission’s AI Act regulatory framework overview. |
| 2 August 2026 | Applicable prohibitions, transparency requirements, and general-purpose AI rules are enforceable, according to the Commission’s AI Act Service Desk. |
| 2 December 2026 | Certain newly amended prohibitions concerning non-consensual intimate material and child sexual abuse material apply, according to the Commission’s AI Act Service Desk. |
| 2 December 2027 | Rules for Annex III high-risk AI systems apply, according to the Commission’s AI Act Service Desk. |
| 2 August 2028 | Rules for high-risk AI systems embedded in regulated products apply, according to the Commission’s AI Act Service Desk. |
Application and enforceability are not interchangeable terms, and the timetable is staged rather than a single start date. The Commission’s official pages are the appropriate references for the EU schedule; dates can change as rules are amended or implementation details evolve.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.How to assess the argument
The practical question is not simply “lawsuits or regulation?” It is which mechanism addresses which gap. For a particular AI-related harm, readers can ask:
Free tools Windows power users keep installed
One-click scans. No signup required.
- What remedy is needed after harm? Consider whether an affected person can bring a claim under applicable law and what remedy might be available.
- What should be required in advance? Identify whether a defined duty, transparency rule, or restriction could reduce a foreseeable risk before a dispute arises.
- Who and what are covered? Check the relevant jurisdiction, the actor’s role, and whether the system falls within the law’s scope rather than assuming every AI system is treated alike.
- Who enforces the rule? Consider whether claimants can realistically pursue litigation or whether the regulator has the authority and capacity to oversee compliance.
This framework does not determine which policy is best in every case. It clarifies why one remedy need not make the other redundant: liability can address responsibility for damage, while regulation can set prospective requirements. Marcus’s essay is a case for rejecting the either-or framing, not a comprehensive legal guide.
Best Value
Further reading on Marcus’s argument
Marcus’s book Taming Silicon Valley is identified in the essay as an earlier treatment of technology-company accountability. It may provide context for his broader position, but it is not current legal guidance.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
