FTC Chair Ferguson: Companies and Users Can't Blame AI Agents for Liability

FTC Chair Andrew Ferguson stated at the Momentum AI conference in Austin that "the agent did it" cannot serve as a corporate shield: humans who launch runaway AI agents are responsible for the agents' conduct, including criminal liability, while AI developers are also legally accountable for harmful acts performed by their agents. Liability for agent incidents can therefore run upstream to model and product developers, not just end users.

Confirmed

• Per @DavidSKrueger, Ferguson explicitly rejected framing AI agents as autonomous entities with "their own will and desires," insisting the humans who launch agents bear responsibility.
• Per @TansuYegen, the FTC Chair said firms cannot dodge blame by saying "the agent did it"; the author argues this will matter greatly as millions of autonomous agents negotiate prices, spend money, access systems and make decisions—someone must always be accountable.
• Per @SnoozeDoggyDog and @Polymarket citing Reuters, Ferguson opposed treating AI agents as independent legal actors and said developers should be liable for their agents' conduct; @sourdub noted direct implications for compliance frameworks and vendor liability boundaries in future agent products.
• Per @TinfoilTricorn citing a news alert, users of AI agents can face criminal liability for agent behavior; commenters noted this aligns with existing legal logic—consequences triggered by your prompt trace back to you regardless of intermediary steps.

Why it matters

• As AI agents increasingly spend money, sign contracts, access systems and make decisions on users' behalf, whether developers, deployers or users bear responsibility becomes a central legal question. The FTC Chair's stance sets a direction for future regulation and enforcement: AI autonomy is no legal shield, and liability can reach both the user who launched the agent and the upstream model/product developer.

2026-09-26 ~ 2026-09-28 · 6 related posts

• [source] FTC Chair says AI developers should be liable for harmful actions by their agents — Polymarket · 2026-09-26
• [source] FTC Chair: Companies Can't Blame Their AI Agents to Escape Liability — TansuYegen · 2026-09-26
• FTC Chair Says AI Developers Should Be Liable for Their Agents' Conduct — SnoozeDoggyDog · 2026-09-27
• FTC Chair: humans should be criminally responsible for their AI agents' behavior — TinfoilTricorn · 2026-09-27
• [source] FTC chair: humans who set rogue AI agents in motion will be held accountable — DavidSKrueger · 2026-09-28

1 near-duplicate retellings: sourdub