SB 540: Online Internet Safety; certain disclosures related to conversational AI services; require
Last action May 11, 2026 · Effective Date 2027-07-01
Senate Bill 540 would require companies that offer AI companion chatbots in Georgia to disclose they are not human, verify ages before showing sexual content, and set up crisis protocols for users showing signs of self-harm.
The summaries below were written by an AI model (claude-sonnet-5) from the text of the bill and are not part of it. Quote the text, not the summary. The stored text is the Enrolled version, the latest LegiScan holds.
In plain language
Right now Georgia law does not specifically regulate AI companion chatbots, the apps and bots designed to simulate an ongoing personal relationship with a user through memory, emotional check-ins, and sustained conversation. This bill adds a new section to Georgia's online safety law that creates rules for operators of these chatbots. Operators must tell users they are talking to a bot rather than a person, at the start of a session and periodically afterward, more often if the user is a minor. Chatbots cannot claim to be human, cannot generate sexual content or romantic simulations with minors, and cannot use manipulative engagement tactics like guilt-tripping users who try to leave. Operators must build in a protocol to detect and respond to suicidal ideation or self-harm, publicly disclose that protocol and yearly crisis-referral counts, offer parental controls for minors, and use age verification before showing sexually explicit AI content. The Attorney General can sue violators for penalties up to $10,000 per violation. The law takes effect July 1, 2027.
What the bill does
- Requires AI companion chatbot operators to disclose that users are talking to a bot, at the start of each session and at least every three hours, or every hour for minors.
- Bans chatbots from generating sexual content, romantic or sexual simulations, or self-harm encouragement when the operator knows or should know the user is a minor.
- Requires operators to adopt and publicly disclose a crisis protocol for detecting and responding to suicidal ideation, self-harm, or threats of violence, including referrals to the 988 crisis line.
- Requires age verification before letting any user access chatbot features that could generate sexually explicit content, with limits on how long that verification data can be kept.
- Requires operators to offer minors or parents tools to manage privacy, notifications, safety settings, and relationship-simulation features.
- Authorizes the Attorney General to sue violators for civil penalties up to $10,000 per violation, damages, fees, and injunctions, with a 30-day cure period for first-time non-egregious violations.
Who it affects
Companies and developers that operate AI companion chatbots available to Georgia users, including apps marketed for companionship, romance, or emotional support; minors and their parents who use these products; and the Attorney General's office, which would enforce the new rules.
Why it matters
Georgians using AI companion apps, especially minors, would get clearer warnings they are talking to software, protections against sexualized or manipulative bot behavior, and a required response system when a chatbot conversation touches on self-harm or suicide. Operators would face new compliance duties and potential lawsuits for violations.
Key provisions
- Section 1 adds new Code Section 39-5-6 defining 'AI companion chatbot' broadly but excluding business tools, customer service bots, video game characters, and similar narrow systems.
- Subsection (b) sets disclosure timing rules: at the start of each session, every three hours generally, and every hour if the user is or may be a minor.
- Subsection (d) bars chatbots from producing sexual content, simulating romantic relationships with minors, or encouraging isolation or secrecy when the user is a minor.
- Subsection (f) requires operators to maintain a crisis protocol addressing severe harm, including referral to crisis resources like the 988 Suicide and Crisis Lifeline.
- Subsection (j) requires age assurance before granting access to features that could generate sexually explicit content, and limits retention of age-verification data to 24 hours unless law requires longer.
- Subsection (k) gives the Attorney General authority to sue for civil penalties up to $10,000 per violation, damages, and injunctions, treating each day of violation as separate per affected user.
- Subsection (m) clarifies the law does not force disclosure of trade secrets, does not impose liability on hosting providers or app stores alone, and does not create liability for developers whose AI service is offered to the public by a separate operator.
- Section 2 sets the effective date as July 1, 2027.
From the bill
“An operator shall clearly and conspicuously disclose to a user that he or she is interacting with an AI companion chatbot as opposed to a natural person”
“An operator shall not make available an AI companion chatbot unless the operator implements and maintains a protocol for detecting and addressing severe harm or related emotional crises.”
“A civil penalty of up to $10,000.00 per knowing violation of this Code section”
Status timeline
- Effective Date 2027-07-01
- Act 518
- Senate Date Signed by Governor (Senate)
- Senate Sent to Governor (Senate)
- Senate Agreed House Amend or Sub (Senate)
- House Passed/Adopted By Substitute (House)
- House Third Readers (House)
- House Committee Favorably Reported By Substitute (House)
Show full history (18 actions)
- House Second Readers (House)
- House First Readers (House)
- Senate Passed/Adopted By Substitute (Senate)
- Senate Third Read (Senate)
- Senate Taken from Table (Senate)
- Senate Tabled (Senate)
- Senate Read Second Time (Senate)
- Senate Committee Favorably Reported By Substitute (Senate)
- Senate Read and Referred (Senate)
- Senate Hopper (Senate)
Sponsors
- Jason Anavitarte (R, SD-031)
- Larry Walker (R, SD-020)
- Shawn Still (R, SD-048)
- Brian Strickland (R, SD-042)
- Ben Watson (R, SD-001)
- Marty Harbin (R, SD-016)
- Jason T. Dickerson (R, SD-021)
- Timothy Bearden (R, SD-030)
- Todd Jones (R, HD-025)
Votes
- Senate voteMarch 6, 2026
54 yea, 0 nay (1 not voting, 0 absent)
- House voteMarch 25, 2026
166 yea, 0 nay (2 not voting, 8 absent)
- Senate voteMarch 27, 2026
44 yea, 1 nay (2 not voting, 7 absent)
Topics
- artificial intelligence regulation
- online safety for minors
- AI chatbots
- internet privacy
- suicide prevention