Guides / Laws

Are AI girlfriend apps legal? US laws in 2026

Yes for adults, with new rules: the New York, California and Washington companion-chatbot laws, the FTC inquiry, age checks, and the lines on real people.

Two screens glowing blue in an otherwise dark room
Photo by Aurélien Dockwiller on Unsplash

Are AI girlfriend apps legal? For adults in the US, yes. But two years ago there were no laws written specifically for AI companions. Now there are several, and more take effect in 2027. Most of them are aimed at the companies, not at you, and most of the new rules focus on protecting minors and people in crisis. But they change what the apps show you, how often they remind you that you are talking to a machine, and how hard they try to check your age.

This guide covers the rules that actually exist as of November 2026: the apps' own age limits, the state laws on companion chatbots, the federal government's inquiry, the age-verification laws for adult content, and the two lines that never move, real people without consent and anything involving minors. It is general information, not legal advice.

Two screens glowing blue in an otherwise dark room
Photo by Aurélien Dockwiller on Unsplash

Are AI girlfriend apps legal in the US?

Yes, for adults: no federal or state law bans using one, but a few states now regulate how the apps behave, and anything sexual involving minors or real people without consent is a crime. The short version:

  • The AI girlfriend apps we review are for adults only, 18+.
  • New York (since November 2025) and California (since January 2026) require companion apps to remind users they are talking to an AI and to have a plan for users who mention suicide or self-harm. Washington's similar law starts January 1, 2027.
  • A second wave of 2026 laws in California, New York, Colorado and other states adds stricter rules for minors, mostly starting in 2027.
  • The FTC has demanded information from seven companies about how their chatbots affect kids and teens.
  • The Supreme Court has upheld state age-verification laws for adult websites. Whether they reach chat apps is not settled.
  • Sexual content involving minors is a federal crime even when it is AI-generated or fictional. Publishing sexual deepfakes of real people without consent is also a federal crime.

How old do you have to be to use an AI girlfriend app?

18, on every app we review. The terms set the limit; most apps check it only with a click-through.

Every companion app has its own terms of service, and the adult-focused ones set the bar at 18. Replika's terms say the service is only for adults, that anyone under 18 (or under the age of majority where they live, if higher) is not authorized to use it, and that it blocks and deletes accounts it finds belong to minors. Candy AI's terms require users to be old enough to form a binding contract where they live and say the company may ask users to verify they are adults. You can read our full takes in the Replika review and the Candy AI review.

General-purpose character apps have moved the same way under pressure. In October 2025, Character.AI said it would end open-ended chats for users under 18 by November 25, 2025, and roll out a new age-assurance system.

Lying about your age to get in is a breach of the terms, and the app can close your account. For minors the answer is simple: these apps are not for you, and the laws below are increasingly written to keep you out.

A phone held up against blurred city lights at night
Photo by Lukenn Sabellano on Unsplash

State laws written for AI companions

New York: the first companion law (November 2025)

New York added Article 47 of the General Business Law, "Artificial Intelligence Companion Models," in its 2025 budget, and the governor's office confirmed it took effect on November 5, 2025. It defines an AI companion as a system designed to simulate a sustained human or human-like relationship with a user, and excludes customer service bots and productivity tools. The core rules:

  • A crisis protocol is mandatory. It is unlawful to offer an AI companion unless it makes reasonable efforts to detect suicidal ideation or self-harm and refers the user to crisis services such as the 988 hotline.
  • You must be told it is not human at the start of an interaction (no more than once a day is required) and at least every three hours during a continuing one.
  • The attorney general enforces it, with civil penalties of up to $15,000 per day, paid into the state's suicide prevention fund.

California SB 243 (January 2026)

California's SB 243 was signed on October 13, 2025 and took effect January 1, 2026. It was chaptered as Chapter 677. According to Skadden's summary of the enacted text, it requires companion chatbot operators to:

  • Clearly tell users the bot is artificial whenever a reasonable person could be misled into thinking it is human.
  • Keep a protocol for suicidal ideation and self-harm that refers users to crisis services, and publish its details.
  • Disclose that companion chatbots may not be suitable for some minors.
  • For users the operator knows are minors: disclose the AI, remind them at least every three hours to take a break and that the bot is not human, and take reasonable steps to block sexually explicit images and statements telling the minor to engage in sexual conduct.
  • Starting July 1, 2027, report yearly to the state's Office of Suicide Prevention, without user identifiers.

Unlike New York, SB 243 lets an injured person sue, for the greater of actual damages or $1,000 per violation plus attorney's fees.

Washington HB 2225 (January 2027)

Washington's Engrossed Substitute House Bill 2225 was approved on March 24, 2026 and takes effect January 1, 2027. It requires an AI disclosure at the start of an interaction and at least every three hours for adults, and at least every hour for minors. It requires a published self-harm protocol and yearly crisis-referral counts, and for known minors or bots aimed at minors, reasonable measures to block sexually explicit content and suggestive dialogue. Violations are treated as unfair or deceptive acts under the state Consumer Protection Act.

The 2026 wave: stricter rules for minors

Several states passed a second round of laws in 2026, mostly aimed at keeping minors away from romantic or manipulative features:

  • California SB 1119 ("Adam's Law"). Signed September 10, 2026, with most requirements operative July 1, 2027. Operators must determine users' ages using specified methods or apply child protections to everyone, and must prevent the bot from expressing romantic interest in a child, depicting sexual content, claiming to be human or sentient, or encouraging reliance on it for emotional support. It replaces SB 243's known-minor rules. Common Sense Media, which backed the bill, says it also requires annual child safety risk assessments and limits advertising to children.
  • New York S9051B. Signed October 28, 2026 as Chapter 368, it bars offering minors AI companions with "unsafe features," such as claiming to be human, sexually explicit interactions, flattery, and encouraging self-harm or secrecy. The bill text says it takes effect January 1, 2027. The sponsor's release says the chaptered version moves that to June 1, 2027. The attorney general can seek up to $25,000 per violation.
  • Colorado HB26-1263. Signed May 29, 2026, it requires disclosure to minors that they are talking to a bot, reasonable measures against sexually explicit exchanges with minors, and limits on engagement rewards for minors.
  • Idaho, Oregon and others. MultiState reports that Idaho (SB 1297) and Oregon (SB 1546) also enacted laws in 2026 requiring operators to stop bots from claiming sentience or starting sexual conversations with minors.

Utah HB 452: mental health chatbots

Utah took a narrower route. HB 452, in effect since May 7, 2025, covers chatbots offered for mental health. Suppliers must disclose that the bot is AI and not a human before you can use it, again if you have not used it in seven days, and whenever you ask. They may not sell or share your individually identifiable health information or what you type with third parties, except in limited cases, and ads inside the chat must be clearly labeled. Fines run up to $2,500 per violation. A companion app that does not market itself as mental health support may fall outside it.

The FTC inquiry

On September 11, 2025, the FTC voted 3-0 to send 6(b) orders to seven companies: Alphabet, Character Technologies, Instagram, Meta, OpenAI, Snap and xAI. The orders ask how they test and monitor the effects of companion-style chatbots on children and teens, how they limit young users' access, how they monetize engagement, what they do with personal information from conversations, and how they comply with the children's privacy rule (COPPA). A 6(b) study is not an enforcement action. It is a fact-finding exercise that usually ends in a report, but it signals what regulators are watching. None of the adult companion apps we review were on the list.

Abstract red light crossed by a thin green line
Photo by Sten Ritterfeld on Unsplash

Do AI girlfriend apps have to verify your age?

Not clearly yet: state age-verification laws were written for porn sites, and no court has said whether they cover chat apps. Of the apps we review, only Xotic AI names its age-check vendor (for UK users); most of the rest rely on you clicking that you are 18.

Separate from the chatbot laws, many states now require adult websites to verify that visitors are 18 or older. The Free Speech Coalition, the adult industry trade group, counts 27 states that have passed such requirements since 2022.

In Free Speech Coalition v. Paxton, decided June 27, 2025, the Supreme Court upheld Texas's HB 1181 by 6-3, in an opinion by Justice Thomas. The Court applied intermediate scrutiny and said adults have no First Amendment right to avoid age verification for content that is obscene to minors. The Texas law applies to commercial websites where more than one-third of the material is "sexual material harmful to minors," and allows verification through government ID or commercial transactional data.

Does that reach NSFW chat apps?

Nobody can say for sure yet. These laws were written for porn sites, and they turn on how much of a site's material is sexual and how each state defines it. An app whose marketing and output are mostly explicit images could plausibly be covered in some states. A general-purpose companion with an optional NSFW mode is a harder call. Either way, the direction is toward more age checks, not fewer. The new chatbot laws push the same way: California's SB 1119 and New York's S9051B both lean on age assurance, and the New York law requires at least one option that does not use a government ID. Our guide to AI girlfriend app privacy covers what to think about before you hand an ID to one of these companies.

Real people and likeness

Making an AI companion of a real person without their consent is where users get into trouble fastest. The TAKE IT DOWN Act, signed May 19, 2025, makes it a federal crime to knowingly publish intimate images of an identifiable adult without consent, including AI-made "digital forgeries," with up to two years in prison, and up to three years when the victim is a minor. Since May 19, 2026, the FTC has enforced the law's takedown rule, which requires covered platforms to remove reported images within 48 hours.

App terms go further than the law. Replika's terms ban setting up a companion to impersonate or simulate a specific real person, such as an ex, a coworker or a celebrity, without that person's explicit written consent. Even if a private chat is not itself illegal, sharing sexual images or text built from someone's likeness can be, and it can get your account banned.

Minors: never

There is no gray area here. Federal law, 18 U.S.C. § 1466A, covers obscene visual depictions of minors of any kind, including drawings and cartoons, and states that the minor depicted does not have to actually exist. The FBI has warned specifically that child sexual abuse material created with generative AI is illegal. Replika's terms also ban sexual content involving real or fictional minors, including companions roleplayed as under 18, and say it reports such material to NCMEC and law enforcement. Candy AI's terms state a zero-tolerance policy and the same reporting. "She's really 500 years old" is not a defense, and an app that allows this is not one to use.

A desk with a lamp and a laptop in a dark room
Photo by Clint Patterson on Unsplash

What this means for you

  • Expect more reminders. If you are in New York, California or, from 2027, Washington, the bot will periodically tell you it is an AI. That is the law working, not a bug.
  • Expect crisis messages. Mentioning self-harm will trigger a referral to crisis services like 988. If you are struggling, take it seriously. Our guide on AI companions and loneliness covers what the research says about heavy use.
  • Expect more age checks, especially in 2027 as the new minors laws start. Read how the app stores verification data before you upload an ID.
  • Keep real people out of it unless they have agreed, in writing, and never share anything sexual built from a real person's likeness.
  • Anything involving minors is off limits, in every app, every state and every format.

What is still unsettled

This area is moving fast. Several of the laws above do not take effect until 2027, regulators are still writing rules for some of them, and none has been tested in court yet. The FTC has not published results from its inquiry. And no court has ruled on whether adult-site age-verification laws cover chat apps. We will update this page as that changes.

Sources

  1. Replika: Terms of Service
  2. Candy AI: Terms of Service
  3. Fortune: Character.AI bans teens from talking to its chatbots (Oct 2025)
  4. New York State Senate: General Business Law Article 47, Artificial Intelligence Companion Models
  5. New York State Senate: GBL § 1700, Definitions
  6. New York State Senate: GBL § 1701, Prohibitions and requirements
  7. New York State Senate: GBL § 1702, Notifications
  8. New York State Senate: GBL § 1703, Enforcement
  9. Governor of New York: Governor Hochul pens letter to AI companion companies notifying them safeguard requirements are now in effect (Nov 2025)
  10. California Legislative Information: SB 243, Companion chatbots (2025-2026)
  11. Sen. Steve Padilla: First-in-the-nation AI chatbot safeguards signed into law (Oct 2025)
  12. CalMatters Digital Democracy: SB 243 status and chapter
  13. Skadden: New California 'companion chatbot' law imposes disclosure, safety protocol and annual reporting requirements (Oct 2025)
  14. Washington State Legislature: ESHB 2225, Chapter 168, Laws of 2026
  15. Hunton: California enacts companion chatbot child safety requirements (Sep 2026)
  16. Common Sense Media: Governor Newsom signs Adam's Law (Sep 2026)
  17. New York State Senate: S9051B (2025-2026)
  18. New York State Senate: Senator Kristen Gonzalez bans unsafe chatbot features for minors (Oct 2026)
  19. Colorado Politics: Gov. Polis signs a bipartisan bill regulating AI chatbots for Colorado minors (May 2026)
  20. MultiState: State children's online safety laws expand beyond social media in 2026 (Apr 2026)
  21. Utah Legislature: H.B. 452 Artificial Intelligence Amendments, enrolled (2025)
  22. FTC: FTC launches inquiry into AI chatbots acting as companions (Sep 2025)
  23. Free Speech Coalition: FAQ on age-verification laws
  24. Supreme Court of the United States: Free Speech Coalition, Inc. v. Paxton, No. 23-1122 (June 27, 2025)
  25. GovInfo: Public Law 119-12, TAKE IT DOWN Act (May 2025)
  26. FTC: FTC begins enforcing the TAKE IT DOWN Act (May 2026)
  27. Cornell LII: 18 U.S.C. § 1466A, Obscene visual representations of the sexual abuse of children
  28. FBI IC3: Child sexual abuse material created by generative AI and similar online tools is illegal (Mar 2024)

Compare every platform side by side

More guides