FTC bans DoNotPay’s ‘AI lawyer’ claims and orders refunds

DoNotPay, a company that marketed its online service as “the world’s first robot lawyer,” has faced regulatory action from the Federal Trade Commission (FTC) over misleading artificial intelligence claims. The FTC’s investigation revealed that DoNotPay made unsubstantiated claims about its AI chatbot’s ability to match human lawyer expertise in generating legal documents and providing legal advice.

Key enforcement actions: The FTC has finalized an order requiring DoNotPay to cease making deceptive claims about its AI capabilities and implement significant remedial measures.

• The company must pay $193,000 in monetary relief
• DoNotPay is required to notify all subscribers from 2021-2023 about the FTC settlement
• The order prohibits future advertising claims comparing their service to human lawyers unless backed by substantial evidence

Investigation findings: The FTC’s complaint unveiled significant gaps between DoNotPay’s marketing claims and actual service capabilities.

• The company failed to conduct testing to verify if its AI chatbot performed at the level of a human lawyer
• No attorneys were hired or retained to evaluate the quality and accuracy of the service’s legal features
• The investigation determined that claims of the service being an adequate substitute for human legal expertise were unfounded

Regulatory process: The FTC’s final decision reflects a thorough review process and unanimous agreement among commissioners.

• The Commission received and considered five public comments before finalizing the order
• Commissioners voted 5-0 on January 16, 2025, to approve the final order
• The agency also prepared responses to address the submitted comments

Looking ahead: AI marketing accountability: This enforcement action signals increasing regulatory scrutiny of AI-related marketing claims, particularly in professional services where consumer protection is paramount. Companies developing AI solutions will need to ensure their marketing claims are supported by concrete evidence and testing, especially when comparing AI capabilities to human expertise.