Trade Regulation Rule on Impersonation of Government and Businesses
This final rule prohibits the impersonation of government, businesses, and their officials or agents in interstate commerce. This document contains the text of the final rule and the rule's Statement of Basis and Purpose ("SBP"), including a Regulatory Analysis.
What this rule actually says
The FTC just banned pretending to be someone or something you're not in order to scam people or get them to do something they wouldn't otherwise do. This includes impersonating government agencies (like the IRS), real businesses, or their employees—whether that's via email, chatbot, or any other channel. The rule is specifically about deceptive impersonation in commerce, not satire or parody.
Who it applies to
- If you're building a chatbot that poses as a real company or government agency (even jokingly in a demo), this applies to you.
- If your AI medical scribe, hiring assistant, or support bot clearly identifies itself as AI, you're fine—this rule doesn't cover honest disclosure.
- If you're in the US and engaging in interstate commerce (basically any online business or anything crossing state lines), this applies. International founders targeting US customers: yes, this applies to you too.
- If your product impersonates to gain user trust, extract information, or influence purchasing decisions, you're in violation territory.
- User data scope: The rule cares about *intent to deceive in commerce*—the data you collect matters less than whether the impersonation was the mechanism of deception.
What founders need to do
- Audit your product's identity (1-2 days). Does your AI clearly state it's AI? Does it claim to be a specific company, government agency, or person when it isn't? If yes, fix it now.
- Document your disclosure practices (1 day). Keep records of how you tell users what they're interacting with. This becomes your defense if questions arise.
- Update your terms and any marketing copy (2-3 days). Strip out any language that could be read as impersonation, even if unintentional. Phrases like "I'm your personal assistant from [real company name]" are risky.
- If you're already live, do a content audit (ongoing). Make sure generated outputs (like email templates or chatbot responses) aren't impersonating anyone without explicit user awareness.
- Monitor updates (15 minutes/month). The FTC may issue guidance. Check their website quarterly.
Bottom line
If your AI is honest about being AI and doesn't pretend to be someone else to trick users, monitor this but don't panic—if you're being deceptive about identity in order to gain compliance or sales, act now.