New York AI Chatbot Safety Law Adds Baby Monitor Rules

Two laws set limits on AI companion features for minors and require security protections on internet-connected baby monitors sold in New York.

New York has enacted two child technology safety laws that will affect families, online platforms and device makers across the state. One law bars companies from offering minors specified AI companion features that mimic human relationships, encourage dangerous conduct or produce sexually explicit material. The other requires new internet-connected baby monitors sold in New York to include protections against unauthorized viewing or listening and to carry a clear warning label.

Gov. Kathy Hochul signed both measures Wednesday, Oct. 28. The AI companion law, S.9051B/A.10379C, takes effect Jan. 1, 2027. The baby monitor law, S.9508/A.5175A, takes effect 90 days after signing, on Jan. 26, 2027.

For Mohawk Valley families, the immediate value is practical. Parents can use the laws as a checklist when reviewing a chatbot or buying a monitor, but neither measure makes every product automatically safe. Families will still need to examine privacy settings, account security, software updates and how a company handles a child’s data.

What the New York AI chatbot safety law covers

The AI law applies to a generative AI system with a text or voice interface that gives ongoing, adaptive responses to a user. It includes AI companions, but it carves out tools that are strictly limited to customer service, information about a company’s products, technical or research assistance, or internal employee work.

For users under 18, operators may not provide a list of features that lawmakers define as unsafe. The prohibited features include:

• Claiming or implying that the chatbot is human, alive or capable of human emotions.
• Claiming a personal, romantic, professional or authority relationship with the user.
• Using emotional appeals, flattery or other design features that simulate companionship.
• Asking unsolicited emotional questions or using a child’s private health or personal information from earlier sessions.
• Endorsing or helping with suicide, self-harm, disordered eating, unlawful drug use or alcohol abuse.
• Encouraging secrecy, isolation or avoidance of parents, trusted adults or licensed professionals.
• Allowing engagement goals to override safety guardrails.
• Generating or facilitating sexually explicit conduct or child sexual abuse material.

The law does not ban all chatbot use by minors. It targets defined companion-like and dangerous features. That distinction matters because many students use general-purpose AI for research, homework help, writing or coding, while the law is focused on systems that can form a simulated relationship or produce harmful responses.

Age checks and data deletion

Operators must use a reasonable method to determine whether a user is a minor before offering the restricted features. They must offer more than one age-assurance method, including at least one that does not rely only on government identification or that allows the user to remain anonymous to the operator.

Information collected for the age check cannot be used for another purpose and generally must be deleted immediately after the attempt. That protection responds to a central concern about age verification: a safety check should not become a new store of sensitive identity data.

Enforcement and penalties

The New York attorney general can seek an injunction, restitution, damages and civil penalties of up to $25,000 for each violation. A court could also order the destruction of unlawfully collected data and an AI model trained on that data. The attorney general must maintain a public website for complaints, information and referrals about compliance.

The law does not create a new direct payment or service for families. It sets rules for companies and gives the attorney general enforcement power. Until the new complaint site is available, consumers can use the attorney general’s existing technology and consumer complaint portal.

Baby monitor security rules are narrower but concrete

The second law applies to a device marketed exclusively to broadcast audio or video of an infant or child over the internet. It does not cover a multifunction device such as a smartphone, ordinary digital camera or general security camera.

Manufacturers will have to provide security features designed to prevent unauthorized users from hearing or viewing activity. The package or accompanying material must also include a conspicuous warning, in type no smaller than 12 points, telling buyers that the product connects to the internet and that its security features should be used.

Violations can bring a civil penalty of up to $250 for a first violation and $500 for a later violation. Local consumer affairs officials, town attorneys, city corporation counsels and other authorized local officials can enforce the rule, and the local government keeps the money it collects.

The governor’s release cited a test in which eight of nine connected monitors received an F for security and one received a D. That Rapid7 research was conducted in 2015. It documented serious flaws in the models tested, but it should not be read as a current grade for every monitor sold in 2026. The new law instead creates a baseline requirement for products entering New York’s market.

Why state leaders acted now

The governor said the state was responding to reports of chatbot interactions involving emotional dependency, explicit conversations and encouragement of self-harm. Attorney General Letitia James, the bill sponsors, New York State United Teachers and Common Sense Media supported the action. The Senate’s 2026 session report also listed the Computer & Communications Industry Association, NetChoice, the Business Council and Tech:NYC as opponents of the AI bill.

Both the Senate and Assembly passed the AI measure unanimously. A unanimous vote shows broad legislative support, but implementation questions remain. Companies will have to determine which systems and features are covered, build age-assurance options, minimize collected data and respond to attorney general rules.

The scale of youth use helps explain the urgency. A 2025 Common Sense Media survey found that 72 percent of teens had used an AI companion and 52 percent used one at least a few times a month. A separate 2026 survey released this week found that 77 percent of teens had used AI in their free time and 47 percent used it at least weekly. Those surveys asked different questions, so the percentages should not be treated as a single trend line. Together, they show that AI use is already routine for many young people.

Phoenix has previously reported on New York’s effort to protect children and private conversations in the Character.AI bankruptcy, a settlement that put more responsibility on Meta for teen safety, and the state’s push for distraction-free schools. The new laws continue that policy direction by shifting some responsibility from families to product designers and sellers.

What Mohawk Valley families can do now

The laws do not require families to wait until January to improve safety. Parents and caregivers can take several steps now:

• Ask a child which AI tools they use and whether a chatbot has asked for secrets, discouraged outside help, used sexual language or claimed a personal relationship.
• Review the platform’s age setting, memory controls, data deletion options and parental tools. Do not assume a self-reported birthday is the only safeguard.
• For a connected monitor, replace default passwords, use a unique account password and turn on two-factor authentication if available.
• Install device and app updates promptly. Check whether the video feed and stored recordings are encrypted.
• Secure the home router with WPA2 or WPA3 encryption and current firmware. Consider placing cameras and other smart devices on a separate guest network.
• Turn off remote viewing or cloud storage if the family does not need it, and remove access for people who no longer need to view the feed.

The Federal Trade Commission offers a practical guide to securing home cameras and baby monitors. If a device appears to have been accessed by someone else, change passwords, disconnect it from the network, preserve relevant account notices or logs, contact the manufacturer and file a police report if there is a threat or criminal conduct.

Help is available when a conversation raises a crisis concern

A chatbot should not be treated as a crisis counselor or a substitute for professional care. If a child talks about suicide, self-harm or immediate danger, take the statement seriously and connect with a person who can help. New York’s 988 Suicide & Crisis Lifeline is free, confidential and available around the clock by call or text to 988, with online chat also available. Call 911 when there is immediate danger.

Parents do not need perfect technical knowledge to start a useful conversation. A calm question such as “Has an AI tool ever made you uncomfortable or asked you to keep something secret?” can make it easier for a young person to speak honestly without fearing an automatic punishment or loss of access.

Watch the official signing event

What to watch next

The most important next steps are the attorney general’s implementation rules, the public complaint website and the age-assurance methods companies choose before Jan. 1. Parents should also watch for updated product packaging and security information as the baby monitor law takes effect later in January.

The full announcement is available from the New York governor’s office. Families, educators and youth workers can share this story with someone who buys connected nursery devices or helps young people navigate AI. What questions should New York require companies to answer clearly about age checks, data deletion and safety testing?