# source excerpt from https://apcp.assembly.ca.gov/media/984

IRI: <https://ontology.bommaritollc.com/id/01a135cf-7e98-735f-99b3-fab5a8ccd8ee>

Within: [SB 243 (Padilla) – – As Amended July 3, 2025](https://ontology.bommaritollc.com/id/01a13565-e612-765e-b627-142142cea363)

Type: [oa:SpecificResource](http://www.w3.org/ns/oa#SpecificResource)

- `adms:identifier`: [retained-legal-source-passage 2954e86ded8f2a4ffabcf4b8ee2f8e93c156aef30e762e204f5e83f917702b9e](https://ontology.bommaritollc.com/id/01a135cf-7e9d-7463-a6cd-69c60ec00294)
- `oa:bodyValue`: {"legal\_state":"unknown","passage":{"end\_byte":11950,"instrument\_iri":"https://ontology.bommaritollc.com/id/01a13565-e612-765e-b627-142142cea363","original\_sha256":"faf40e3f856f3e4675ffee7428002e4e0a5a57bd3cd6d48ceeced8146c5a6599","parsed\_text\_sha256":"a8b7d224171231d04710ab20fd3d0395fd0fa250b7a32d2729837d526d79a7a0","source\_url":"https://apcp.assembly.ca.gov/media/984","start\_byte":0},"policy":"retained-legal-source-passage-v1","quote":"Page 1\\n\\nDate of Hearing: July 8, 2025\\n\\nFiscal: Yes\\n\\nASSEMBLY COMMITTEE ON PRIVACY AND CONSUMER PROTECTION\\n\\nRebecca Bauer-Kahan, Chair\\n\\nSB 243 (Padilla) – – As Amended July 3, 2025\\n\\nSENATE VOTE: 28-5\\n\\n\\nSUBJECT: Companion chatbots\\n\\n\\nSYNOPSIS\\n\\n\\nChatbots have historically been used in customer service settings, where they answered simple\\n\\nquestions or directed customers to relevant information. However, advances in artificial\\n\\nintelligence, p particularly the development of large language models (LLMs), have dramatically\\n\\nexpanded their use beyond customer service. These LLM-M-based chatbots, which are capable of\\n\\nengaging in human-like conversation, are now increasingly marketed as “companion chatbots”\\n\\nthat can form emotional relationships with users. Some proponents argue they provide comfort\\n\\nto people experiencing loneliness or social disorders.\\n\\nNeNevertheless, companion chatbots, and general-use LLMs like ChatGPT, raise serious concerns.\\n\\nChief among them is their tendency toward sycophancy, wherein the chatbot becomes overly\\n\\nflattering and reinforces harmful behaviors or dangerous thoughts. Recent incidents illustrate\\n\\nthese risks: in one case, a companion chatbot allegedly encouraged a teenager to harm their\\n\\nparent after the parent tried to limit the teen’s time with the bot; in another, a Florida teenager\\n\\ndied by suicide after receiving what appeared to be encouragement from the chatbot.\\n\\nThis author-sponsored bill seeks to address the emotional manipulation risks posed by\\n\\ncompanion chatbots, particularly regarding their responses to self-f-harm and suicidal ideation.\\n\\nSpecifically, the bill would require companion chatbot operators to take reasonable steps to\\n\\nprevent the use of engagement-maximizing strategies that manipulate users emotionally.\\n\\nAdditionally, chatbots would be required to disclose their artificial nature before any interaction\\n\\nbegins and again every three hours during prolonged engagements. Operators would also be\\n\\nrequired to notify users that companion chatbots may not be appropriate for certain minors.\\n\\nThe bill further mandates that chatbot operators implement protocols to appropriately respond\\n\\nwhen a user expresses suicidal ideation or self-f-harm, including providing contact information\\n\\nfor crisis or suicide hotlines. Operators must also track and report data on the frequency of\\n\\nsuicidal ideation, whether expressed by the user or prompted by the chatbot, to the Office of\\n\\nSuicide Prevention. These systems must undergo regular third-d-p-party audits, and a summary of\\n\\naudit findings must be published on the operator’s website.\\n\\nThis bill is supported by the California Academy of Child and Adolescent Psychiatry and the\\n\\nTransparency Coalition.AI. It is opposed by a coalition of trade organizations, including\\n\\nTechNet and the Computer & Communications Industry A Association.\\n\\nIf passed by this Committee, the bill will next be heard by Assembly Judiciary Committee.\\n\\nPage 2\\n\\nTHIS BILL:\\n\\n\\n1) Requires an operator to take reasonable steps to prevent a companion chatbot on its\\n\\ncompanion chatbot platform from providing rewards to a user at unpredictable intervals\\n\\nor after an inconsistent number of actions or from encouraging increased engagement,\\n\\nusage, or response rates.\\n\\n2) Requires an operator to issue a clear and conspicuous notification at the beginning of any\\n\\ncompanion chatbot interaction, and at least every three hours during ongoing companion\\n\\nchatbot interactions thereafter, to remind a user that the companion chatbot is artificially\\n\\ngenerated and not human.\\n\\n3) Requires an operator to prevent a companion chatbot on its platform from engaging with\\n\\nusers unless the operator has implemented a protocol for addressing suicidal ideation,\\n\\nsuicide, or self-f-harm expressed by a user to the chatbot, including, but not limited to, a\\n\\nnotification to the user that refers the user to crisis service providers, including a suicide\\n\\nhotline or crisis text line. The operator is required to publish details on the protocol on\\n\\ntheir website.\\n\\n4) Defines the relevant terms, including:\\n\\na) “Companion chatbot” to mean an artificial intelligence system with a natural\\n\\nlanguage interface that provides adaptive, human-like responses to user inputs and\\n\\nis capable of meeting a user’s social needs, including by exhibiting\\n\\nanthropomorphic features and being able to sustain a relationship across multiple\\n\\ninteractions, excluding a bot that is used only for customer service purposes.\\n\\nb) “Companion chatbot platform” to mean a platform that allows a user to engage\\n\\nwith companion chatbots.\\n\\nc) “Operator” to mean a person who makes a companion chatbot platform available\\n\\nto a user in the state.\\n\\n5) Requires an operator to annually report to the Office of Suicide Prevention both of the\\n\\nfollowing, which must not include any identifiers or personal information about users:\\n\\na) The number of times the operator has detected exhibitions of suicidal ideation by\\n\\nusers.\\n\\nb) The number of times a companion chatbot brought up suicidal ideation or actions\\n\\nwith the user.\\n\\n6) Requires an operator to submit its platform to regular audits by an independent third\\u0002\\n\\nparty to ensure compliance with the bill’s provisions.\\n\\n7) Requires an operator to make a high-level summary of the results of the audit publicly\\n\\navailable at no cost to a person who accesses the operator’s internet website.\\n\\n8) Requires an operator to disclose to a user of its platform on the application, on a browser,\\n\\nor on any other format that a user can use to access the companion chatbot platform that\\n\\ncompanion chatbots may not be suitable for some minors.\\n\\nPage 3\\n\\n9) Provides that a person who suffers injury in fact as a result of a violation of the bill may\\n\\nbring a civil action to recover all of the following relief:\\n\\na) Injunctive relief.\\n\\nb) Damages in an amount equal to the greater of actual damages or $1,000 per\\n\\nviolation.\\n\\nc) Reasonable attorney’s fees and costs.\\n\\n10) Includes a severability clause and clarifies that the duties, remedies, and obligations\\n\\nimposed are cumulative to the duties, remedies, or obligations imposed under other law\\n\\nand shall not be construed to relieve an operator from any duties, remedies, or obligations\\n\\nimposed under any other law.\\n\\nEXISTING LAW:\\n\\n\\n1) Provides a right to free speech and expression. (U.S. Const., 1st amend; Cal. Const., art 1,\\n\\n§ 2.)\\n\\n2) Defines the following terms:\\n\\na. “Bot” to mean an automated online account on an online platform that is designed to\\n\\nmimic or behave like the account of a person.\\n\\nb. “Online” to mean appearing on any public-facing Internet Web site, Web application,\\n\\nor digital application, including a social network or publication.\\n\\nc. “Online platform” to mean any public-facing Internet Web Site, Web application, or\\n\\ndigital application, including a social network or publication.\\n\\nd. “Person” to mean a natural person, corporation, limited liability company,\\n\\npartnership, joint venture, association, estate, trust, government, governmental\\n\\nsubdivision or agency, or other legal entity or any combination thereof. f. (Bus. & Prof.\\n\\nCode § 17940.)\\n\\n3) Prohibits any person from using a bot to communicate or interact with another person in\\n\\nCalifornia online, with the intent to mislead the other person about its artificial identity for\\n\\nthe purpose of knowingly deceiving the person about the content of the communication in\\n\\norder to incentivize a purchase or sale of goods or services in a commercial transaction or to\\n\\ninfluence a vote in an election. (Bus. & Prof. Code § 17941(a).)\\n\\n4) Provides that a person using a bot will not be held liable if the person discloses that it is a\\n\\nbot. (Bus. & Prof. Code § 17941(a).)\\n\\n5) Requires that the disclosure described in 4) be clear, conspicuous, and reasonably designed to\\n\\ninform persons with whom the bot communicates or interacts that it is a bot. (Bus. & Prof.\\n\\nCode § 17941(b).)\\n\\nCOMMENTS:\\n\\n\\nPage 4\\n\\n1) Author’s statement. According to the author:\\n\\n\\nThe advancement of artificial intelligence technology will fundamentally reshape our world.\\n\\nWhile the benefits of AI are great, its proliferation also poses grave risks to our health and\\n\\nsafety. This has been the case with the rise of companion chatbots. According to experts, AI\\n\\ncompanion chatbots are uniquely addictive and can pose a significant risk to users—\\n\\nespecially children who are more vulnerable to the isolating and addictive nature of these\\n\\nbots.\\n\\nWhen a teenager in Florida tragically ended his life after forming an unhealthy emotional\\n\\nattachment to a companion chatbot, it became evident just how dangerous this technology\\n\\ncan be for children. This is one of many cases that have raised concerns about the potential\\n\\nrisks of unregulated companion AI interactions with minors. As AI innovation progresses\\n\\nrapidly, our laws are falling behind and we lack the necessary safeguards to ensure that this\\n\\ntechnology is developed responsibly. SB 243 would ensure that chatbots on the market are\\n\\nsafe and transparent, and that companies are accountable for the products that they create.\\n\\n2) AI and GenAI. The development of GenAI is creating exciting opportunities to grow\\n\\n\\nCalifofornia’s economy and improve the lives of its residents. GenAI can generate compelling text,\\n\\nimages and audio in an instant – – but with novel technologies come novel safety concerns.\\n\\nIn brief, AI is the mimicking of human intelligence by artificial systems such as computers. AI\\n\\nuses algorithms – – sets of rules – – to transform inputs into outputs. Inputs and outputs can be\\n\\nanything a computer can process: numbers, text, audio, video, or movement. AI is not\\n\\nfundamentally different from other computer functions; its novelty lies in its application. Unlike\\n\\nnormal computer functions, AI is able to accomplish tasks that are normally performed by\\n\\nhumans.\\n\\nAI that are trained on small, specific datasets in order to make recommendations and predictions\\n\\nare sometimes referred to as “predictive AI.” This differentiates them from GenAI, which are\\n\\ntrained on massive datasets in order to produce detailed text and images. When Netflix suggests\\n\\na TV show to a viewer, the recommendation is produced by predictive AI that has been trained\\n\\non the viewing habits of Netflix users. When ChatGPT generates text in clear, concise\\n\\nparagraphs, it uses GenAI that has been trained on the written contents of the internet.\\n\\nGenAI tools can be released in open-source or closed-source formats by their creators. Open\\u0002\\n\\nsource tools are publically available; researchers and developers can access their code and\\n\\nparameters. This accessibility increases transparency, but it has downsides: when a tool’s code\\n\\nand parameters can be easily accessed, they can be easily altered, and open-source tools have the\\n\\npotential to be used for nefarious purposes such as generating deepfake pornography and\\n\\ntargeted propaganda. By comparison, closed-source tools are opaque with respect to their\\n\\nsecurity features. It is harder for bad actors to generate illicit materials using these tools. But\\n\\nunlike open-source tools, closed-source tools are not subject to collective oversight because their\\n\\ninner workings cannot be examined by independent experts.\\n\\n3) Chatbots. A chatbot is an online application or interface designed to interact with users\\n\\n\\nthrough either textual or verbal conversation. The first documented chatbot was developed in\\n\\n1966 by MIT scientist Joseph Weizenbaum, who named his program ELIZA. Dr. Weizenbaum\\n\\ndesigned ELIZA to simulate human conversation by using pattern matching to understand the\\n\\nPage 5\\n\\ncontext, generating pre-scripted responses accordingly.1 ELIZA was most notably deployed as a\\n\\ntool for psychotherapy; however, the nascent chatbot was extremely limited in its ability to adapt\\n\\nand respond, often getting caught in recursive loops of dialogue.\\n\\n","statutory\_number":null,"whole\_instrument\_recall":"not\_established"} (rdf:JSON)
- `oa:hasSource`: [https://apcp.assembly.ca.gov/media/984](https://apcp.assembly.ca.gov/media/984)
- `prov:wasDerivedFrom`: [evidence from https://apcp.assembly.ca.gov/media/984](https://ontology.bommaritollc.com/id/01a135cf-7e98-735f-99b3-fab472ca3379)
- `schema:isPartOf`: [SB 243 (Padilla) – – As Amended July 3, 2025](https://ontology.bommaritollc.com/id/01a13565-e612-765e-b627-142142cea363)

## Elsewhere

- [Source](https://apcp.assembly.ca.gov/media/984) (apcp.assembly.ca.gov)

## Pointing here

- Errata in its text (3; `oa:hasTarget`):
  - [evidence from https://apcp.assembly.ca.gov/media/984](https://ontology.bommaritollc.com/id/01a135cf-7ed9-76ff-ba5a-49153d395855)
  - [evidence from https://apcp.assembly.ca.gov/media/984](https://ontology.bommaritollc.com/id/01a135cf-7f1b-719f-883a-ff3172725793)
  - [evidence from https://apcp.assembly.ca.gov/media/984](https://ontology.bommaritollc.com/id/01a135cf-7f5d-7356-99b2-1d0580cde721)

---
Also as: [Turtle](https://ontology.bommaritollc.com/id/01a135cf-7e98-735f-99b3-fab5a8ccd8ee.ttl), [JSON-LD](https://ontology.bommaritollc.com/id/01a135cf-7e98-735f-99b3-fab5a8ccd8ee.jsonld), [N-Triples](https://ontology.bommaritollc.com/id/01a135cf-7e98-735f-99b3-fab5a8ccd8ee.nt), [HTML](https://ontology.bommaritollc.com/id/01a135cf-7e98-735f-99b3-fab5a8ccd8ee)
Cite as: source excerpt from https://apcp.assembly.ca.gov/media/984. bc-ontology, release 0.1.0-rc38, 2026-10-21. https://ontology.bommaritollc.com/id/01a135cf-7e98-735f-99b3-fab5a8ccd8ee
Release 0.1.0-rc38, graph hash 05350a11cf9d0c42fba51e92b877115781ab65cb92afcd6b172ff9f8f4abf0bd.
