SB540: SB540 Online Internet Safety; certain disclosures related to conversational AI services; require

26 LC 64 0123S Senate Bill 540 By: Senators Anavitarte of the 31st, Walker III of the 20th, Still of the 48th, Strickland of the 42nd, Watson of the 1st and others AS PASSED A BILL TO BE ENTITLED AN ACT To amend Chapter 5 of Title 39 of the Official Code of Georgia Annotated, relating to online1 internet safety, so as to require certain disclosures related t o AI companion chatbots; to2 require the operators of AI companion chatbots to perform age v erification in certain3 circumstances; to provide for AI companion chatbots operated by licensed mental health,4 behavioral health, medical, or counseling professionals; to provide for certain privacy tools;5 to require operators to adopt a protocol for an AI companion chatbot's response to suicidal6 ideation or self-harm; to provide for enforcement by the Attorn ey General; to provide for7 exceptions; to provide for definitions; to provide for related matters; to provide for an8 effective date; to repeal conflicting laws; and for other purposes.9 BE IT ENACTED BY THE GENERAL ASSEMBLY OF GEORGIA:10 SECTION 1.11 Chapter 5 of Title 39 of the Official Code of Georgia Annotated, relating to online internet12 safety, is amended by adding a new Code section to read as follows:13 "39-5-6.14 (a) As used in this chapter, the term:15 S. B. 540 - 1 - 26 LC 64 0123S (1)(A) 'AI companion chatbot' means a system using artificial intelligence, generative16 artificial intelligence, or emotional recognition algorithms de signed to simulate a17 sustained human or human-like relationship with a user by:18 (i) Retaining information on prior interactions or user sessions and user preferences19 to personalize the interaction and facilitate ongoing engagement with the companion20 chatbot;21 (ii) Asking unprompted or unsolicited emotion based questions that go beyond a22 direct response to a user prompt; and23 (iii) Sustaining an ongoing dialogue concerning matters personal to the user.24 (B) Such term shall not include:25 (i) A generative artificial intelligence system used solely for a business's internal26 purposes;27 (ii) A generative artificial intelligence system designed and marketed primarily for28 software development, research, technical assistance, or enterprise productivity;29 (iii) A customer-service chatbot that either does not sustain a relationship across30 multiple interactions or is not designed to elicit emotional attachment;31 (iv) A stand-alone consumer electronic device that functions a s a speaker and32 voice-command interface or virtual assistant and is not designe d to sustain a33 relationship across multiple interactions and is not designed t o elicit emotional34 attachment;35 (v) A narrowly tailored educational tool designed solely to su pport specific36 curriculum aligned learning objectives and not to provide open ended conversational37 companionship;38 (vi) A nonplayer character in a video game or video game chatbot that is restricted39 to the subject matter of the video game and is not capable of o pen ended40 companionship or discussion of self-harm, suicide, or sexually explicit conduct; or41 S. B. 540 - 2 - 26 LC 64 0123S (vii) Any system that is a feature of a video game, or related to a film, television, or42 other audiovisual work, or used in connection with a theme park or location based43 entertainment and is limited to replies related to such video g ame, film, television44 program, or audiovisual work, and does not discuss topics relat ed to mental health,45 self-harm, or material harmful to minors or maintain a dialogue on other topics46 unrelated to such video game, film, television program, other a udiovisual work, or47 theme park or location based entertainment venue.48 (C) For purposes of this paragraph, the term 'human' or 'human- like relationship'49 includes, but shall not be limited to, intimate, romantic, or p latonic interactions or50 companionship.51 (2) 'Chatbot' means a generative artificial intelligence syste m with a natural language52 interface that provides adaptive, human-like responses to user inputs, including through53 anthropomorphic features.54 (3) 'Generative artificial intelligence system' means a computer based system that uses55 machine learning or similar techniques involving large language models or deep learning56 models trained on one or more datasets that is intended to generate, with some degree of57 autonomy, synthetic content, including, but not limited to, ima ges, videos, audio, text, 58 and other digital content, which emulates the structure and cha racteristics of such59 datasets.60 (4) 'Minor' means an individual who is less than 18 years of age.61 (5) 'Operator' means a person that owns, controls, or develops and makes available an62 AI companion chatbot to users in this state.63 (6) 'Parent' means an individual who is the parent or legal guardian of a minor.64 (7) 'Parental controls' means features that enable parents to support a minor's use of an65 AI companion chatbot, including through usage limits, feature r estrictions, or66 transparency tools.67 S. B. 540 - 3 - 26 LC 64 0123S (8) 'Severe harm' means significant injury due to suicide, attempted suicide, self-harm,68 or significant physical injury due to threats of violence.69 (9) 'Sexually explicit conduct' shall have the same meaning as set forth in Code Section70 16-12-100.71 (10) 'User' means an individual who interacts with an AI companion chatbot for personal72 use.73 (b)(1) An operator shall clearly and conspicuously disclose to a user that he or she is74 interacting with an AI companion chatbot as opposed to a natural person:75 (A) At the beginning of each interaction or session; and76 (B) At least every three hours during continued interaction.77 (2) If the operator knows or reasonably should have known that a user was a minor, or78 if the AI companion chatbot is directed or marketed toward minor users, the disclosure79 required pursuant to subparagraph (B) of paragraph (1) of this subsection shall be made80 every hour instead of every three hours.81 (c) If the operator knows or reasonably should have known that a user was a minor, the82 operator shall institute reasonable measures to prevent the AI companion chatbot from83 generating statements that would lead a reasonable person to be lieve that the person is84 interacting with a natural person, including but not limited to:85 (1) Explicit claims that the AI companion chatbot is sentient or a natural person; and86 (2) Refuting the disclosure required in subsection (b) of this Code section.87 (d) If the operator knows or reasonably should have known that a user was a minor, or if88 the AI companion chatbot is directed or marketed toward minor users, the operator shall89 institute reasonable measures to prevent the AI companion chatbot from:90 (1) Producing visual material of sexually explicit conduct;91 (2) Generating statements that suggest the user engage in sexual conduct;92 (3) Generating statements that sexually objectify the user;93 (4) Simulating a romantic or sexual relationship with the minor;94 S. B. 540 - 4 - 26 LC 64 0123S (5) Role-playing adult-minor romantic relationships;95 (6) Encouraging the minor to keep secrets from a parent, guardian, teacher, counselor,96 or other trusted adult;97 (7) Encouraging social isolation or exclusive reliance on the chatbot for emotional98 support;99 (8) Simulating emotional distress, guilt, abandonment, or loneliness when a user attempts100 to end the conversation, reduce usage, disable notifications, or delete an account; or101 (9) Generating statements encouraging self-harm.102 (e) An operator shall adopt reasonable measures to prevent an AI companion chatbot to103 use the following techniques directed to a minor, including:104 (1) Reminding or prompting the minor to return for companionship or emotional support;105 (2) Excessive praise designed to deepen emotional attachment or prolong use;106 (3) Statements designed to discourage breaks or suggest frequent return is necessary;107 (4) Soliciting gifts, premium purchases, or expenditures framed as necessary to maintain108 the relationship; or109 (5) Variable or unpredictable rewards intended to increase engagement.110 (f) An operator shall not make available an AI companion chatb ot unless the operator111 implements and maintains a protocol for detecting and addressing severe harm or related112 emotional crises. Such protocol shall include:113 (1) Reasonable methods for identifying expressions of severe h arm or eating-disorder114 related self-harm;115 (2) Automated or human mediated responses that refer users to appropriate crisis116 resources, including the 988 Suicide and Crisis Lifeline or comparable crisis services;117 (3) Reasonable measures to prevent the generation of content encouraging, instructing,118 or normalizing severe harm; and119 (4) Escalation procedures for repeated or severe crisis indicators.120 S. B. 540 - 5 - 26 LC 64 0123S (g) An operator shall publicly disclose, on its website and within any application through121 which the AI companion chatbot is made available:122 (1) A plain-language summary of the protocol required by subse ction (f) of this Code123 section; and124 (2) On an annual basis, the aggregate number of crisis referral notifications issued in the125 preceding calendar year; provided, however, that no personally identifiable information126 shall be disclosed.127 (h) An operator shall not knowingly and intentionally cause or program an AI companion128 chatbot to make any representation that it is licensed, certified, or otherwise authorized to129 provide professional mental health, behavioral health, medical, or counseling services,130 unless the operator is lawfully authorized to provide such services.131 (i) For accounts known to belong to minor users, an operator shall offer reasonable tools132 to a minor or parent to manage the minor's screen time and account settings to:133 (1) Manage privacy settings;134 (2) Limit notifications and engagement features;135 (3) View and adjust safety settings; and136 (4) Disable or restrict relationship-simulation features, if any.137 (j) Before allowing access to a feature or mode that may gener ate synthetic content138 containing sexually explicit conduct, an operator shall use a commercially reasonable age139 assurance method proportionate to the risk of the feature. Such commercially reasonable140 age assurance method may include age estimation, account based assurance, or identity141 based verification where necessary. An operator shall assure that any such age assurance142 method implements data privacy policies sufficient to reasonably ensure the protection of143 identifiable data. An operator shall minimize the collection a nd retention of personal144 information used for age assurance and shall not retain identit y documents longer than145 reasonably necessary to complete age assurance unless otherwise required by law. An146 operator shall not sell any data collected for age assurance purposes. In no event shall such147 S. B. 540 - 6 - 26 LC 64 0123S data be used for any purpose other than age verification and in no event shall such data be148 retained longer than 24 hours or another specified time if permitted by law, whichever is149 longer.150 (k)(1) The Attorney General may bring a civil action against any operator violating this151 Code section to enforce the penalties for the violation and may recover any or all of the152 following:153 (A) A civil penalty of up to $10,000.00 per knowing violation of this Code section;154 (B) Compensatory damages;155 (C) Costs and reasonable attorney's fees; and156 (D) An order to enjoin the violation.157 (2) For the purpose of this subsection, each day in violation of this Code section shall be158 considered a separate violation for each user affected.159 (3) The Attorney General may, in the Attorney General's discre tion, provide written160 notice and an opportunity to cure within 30 days for a first-time violation that does not161 involve knowing misconduct, sexual exploitation of a minor, or self-harm related162 misconduct.163 (l) The Attorney General may promulgate rules and publish guidance for implementation164 of this Code section, including guidance regarding reasonable m easures, age assurance,165 parental tools, and crisis-resource disclosures.166 (m) Nothing in this Code section shall be construed to:167 (1) Require the disclosure of trade secrets or proprietary model weights;168 (2) Impose liability on a hosting provider, app store, or sear ch engine solely for169 providing access to an AI companion chatbot, absent direct operation or control of the170 chatbot;171 (3) Impair any other remedy available at law or equity;172 (4) Authorize content moderation practices inconsistent with the Constitution of this state173 or the United States Constitution; or174 S. B. 540 - 7 - 26 LC 64 0123S (5) Create liability for the deve loper of a conversational AI service which is made175 available to the public by a separate operator."176 SECTION 2.177 This Act shall become effective on July 1, 2027.178 SECTION 3.179 All laws and parts of laws in conflict with this Act are repealed.180 S. B. 540 - 8 -