AI Compliance for Liability
Liability is addressed by 50 regulatory updates across 7 jurisdictions and 4 frameworks. This page tracks how regulators worldwide are approaching liability in the context of artificial intelligence.
Framework Requirements for Liability
Regulations Covering Liability
US Federal(32)
SEC charges advisers with false AI claims
The SEC charged two investment advisers for making false and misleading statements about their use of artificial intelligence, confirming that AI-wash in financial marketing and disclosures is an enforcement priority.
FTC orders $1 million penalty over deceptive accessibility-compliance AI claims
The FTC took action against deceptive claims that an AI product could make websites WCAG-compliant, reinforcing that capability claims tied to compliance outcomes must be fully substantiated before they are marketed.
FTC finalizes order against DoNotPay over deceptive AI lawyer claims
The FTC finalized an order on 2025-02-xx/2025-02 against DoNotPay that prohibits deceptive 'AI lawyer' claims and imposes monetary relief and consumer-notice obligations, making unsupported AI marketing an immediate enforcement risk.
FTC enforcement against deceptive AI claims continues
The FTC’s AI enforcement actions and related press releases show that deceptive AI performance claims remain an active Section 5 risk, so marketing and product teams need substantiation before regulators ask for it.
FTC seeks public comment on proposed AI accuracy policy statement
The FTC opened public comment on a proposed policy statement targeting AI accuracy claims, signaling that businesses making AI-performance representations should expect scrutiny over substantiation and deception risk now.
California(7)
California AB 2169: Social media platforms: artificial intelligence models
California AB 2169 was read a second time and amended on April 23, 2026, signaling continued legislative attention to AI models on social media platforms and the need to track platform-specific obligations.
State v. Bailey
The 2026-03-30 state court entry contains no substantive excerpt, so the source does not establish any AI, discrimination, or compliance consequence.
Bryan v. Child Support Enforcement Agency
This state court decision dated 2026-03-30 is listed without any substantive excerpt, so no compliance requirement can be extracted from the source material provided.
In the Interest of W.G., Minor Child
This state court ruling dated 2026-04-01 is identified only by caption, with no factual or legal excerpt to indicate any regulatory relevance.
Patrick Hrdlichka v. Samantha Bengston
The source provides only the case name and a 2026-04-01 decision date, with no substantive legal holding to translate into a compliance update.
Colorado(6)
Bryan Dorsey v. Robert T. Jones
The state court listing gives no substantive outcome, so there is no actionable compliance change in the provided record.
State v. Bailey
The state court listing is caption-only and does not identify any operative holding, so no regulatory action is evident from the supplied text.
Bryan v. Child Support Enforcement Agency
The state court listing is caption-only and does not disclose an operative ruling, so there is no compliance delta in the supplied text.
In the Interest of W.G., Minor Child
The minor-child case entry is purely caption-level in the supplied text, with no discernible regulatory or compliance trigger.
Patrick Hrdlichka v. Samantha Bengston
The supplied state court listing lacks any substantive ruling, leaving no actionable compliance change to implement from the available text.
European Union(2)
CJEU preliminary reference on whether AI software can be a high-risk AI system
A 2026 CJEU preliminary reference asks whether software using AI elements qualifies as a high-risk AI system, which could materially affect classification, oversight, and traceability expectations under the AI Act.
CJEU preliminary reference on whether AI-assisted expert reporting is high-risk under the AI Act
A 2026 CJEU reference asks whether software generating automated outcomes or using AI elements in an expert report should be treated as a high-risk AI system under the AI Act, creating interpretive uncertainty for litigation-support tools.
New York(1)
Weekly digest
Leave your email to get each issue in your inbox. Free, no account required.
We use your email only for the digest. Privacy policy