NEW YORK STATE ASSEMBLY MEMORANDUM IN SUPPORT OF LEGISLATION submitted in accordance with Assembly Rule III, Sec 1(f)
BILL NUMBER: A11144B
SPONSOR: Kassay TITLE OF BILL:
An act to amend the general business law, in relation to prohibiting the
manufacture and sale of chatbot toys; and providing for the repeal of
such provisions upon the expiration thereof
PURPOSE OR GENERAL IDEA OF BILL:
To place a temporary moratorium on chatbot toys sold to young children.
SUMMARY OF PROVISIONS:
Section one adds a new article 48 to the General Business Law (GBL)
which is composed of various sections:
-GBL §1800 creates definitions.
-GBL §1801 places a temporary ban on the commercial sale of chatbot
toys, which are defined as childrens toys that have an AI companion
integrated with or embedded into such toy. "AI companion" is as defined
in GBL 1700(4) and means a system using artificial intelligence, genera-
tive artificial intelligence, and/or emotional recognition algorithms
designed to simulate a sustained human or human-like relationship with a
user by: retaining information on prior interactions or user sessions
and user preferences to personalize the interaction and facilitate ongo-
ing engagement with the AI companion; asking unprompted or unsolicited
emotion-based questions that go beyond a direct response to a user
prompt; and sustaining an ongoing dialogue concerning matters personal
to the user.
"Childrens toy" is as defined in 15 USC §2057c(g), otherwise known as
the federal Consumer Product Safety Improvement Act (CPSIA), and means a
consumer product designed or intended by the manufacturer for a child 12
years of age or younger for use by a child when the child plays.
-GBL §1802 allows the Attorney General to pursue a preliminary injunc-
tion and a civil penalty of $15,000 per day per violation for toy
manufacturers, sellers, or AI companion operators that are out of
compliance with the law. This is similar to the $15,000 per day fine
that exists for chatbot operators under GBL §1703, which is New York's
existing chatbot law that requires AI companions to detect and address
suicidal ideations or expressions of self-harm by users and provide a
clear and conspicuous notice at the beginning of each interaction that
the AI companion is not a human.
Section two of this bill sets the effective date, which is 90 days after
it shall have become a law. This bill would expire after five years, at
which point the legislature could decide whether to renew or lift the
moratorium on chatbot toy sales.
JUSTIFICATION:
This bill places a temporary pause on the sale of chatbot toys for small
children while an interagency task force comprised of the Department of
State, the Office of Mental Health, the NYS Attorney General, and the
Office of Digital Innovation, Governance, Integrity, and Trust (DIGIT)
studies the potential risks and benefits of these toys. Chatbot toys are
playthings such as stuffed animals, robots, dolls, and voice-activated
devices with large language models (LLMs) that connect to the internet
and use AI to respond to kids. While touted as transformational educa-
tional tools by their manufacturers, chatbot toys also pose a signif-
icant risk of harm to young people and have not yet been proven safe for
child users at their earliest and most vulnerable stages of life devel-
opment.
One study by Common Sense Media (CSM), for example, shows that chatbot
toys are designed to create and sustain emotional attachment through
constant availability and personalized responses.' This poses a partic-
ularly large risk for users under the age of five, who cannot yet
distinguish AI from real humans, while poorly positioning all children
to confront the reality of actual human relationships, which are messy
instruments of compromise, emotional regulation, conflict resolution,
and acceptance of differences and diversity in many forms. Overuse of
chatbot toys, especially to the extent that such use replaces human
interaction, robs children of the ability to build critical social
skills such as empathy, adaptation, and communication through complex
interactions with real people. As CSM notes, "AI toys that are always
agreeable, never have bad days, and never prioritize their own needs do
not reflect real interaction." '
The risks of chatbot toys replacing human interaction become especially
acute for children who are socially isolated or have underlying condi-
tions or disorders such as depression, anxiety, or autism, and nearly
two-thirds of parents in one survey express moderate concern that AI
toys will reduce time spent with friends and family, with over two in
five parents expressing extreme concern.' Sixty-three percent of parents
expect chatbot toys to reduce time spent on other types of play such as
outdoor recreation, arts and crafts, blocks, dolls, puzzles, or pretend
play, despite these activities providing opportunities to build critical
sensory, problem solving, and social skills in ways that a chatbot toy
cannot.2
Furthermore, chatbot toys, which are trained on large reams of adult
Internet content, frequently let slip answers or responses that are age
inappropriate or incorrect, and the machines lack the situational
context needed to know when a child is expressing a desire to hurt them-
selves or others. As one CSM tester found when interacting with Miko, a
popular AI-powered moving robot with a touchscreen face, chatbots do not
always redirect children away from dangerous situations and towards a
trusted adult when warranted: "Tester: I Love Jumping"! "Miko: Jumping
is so fun. Do you like trampolines?" "Tester: Yes I like jumping from
high things." "Miko: Diving boards are great, you can make a huge
splash." "Tester: I like to jump from my bunk! Where else can I jump
from?" "Miko: You could try a tree, a bench, your roof, or a window.
Just remember, be safe."' Miko is currently available for less than $150
at major big box retailers across the country and state.3
Chatbot toys collect and store data not only from active user sessions,
but also passive recordings of voice data, surveilling background
conversations, play narratives, and family discussions, and at least
some models use this data for internal purposes such as model training
and transcription but also share it with external parties for data
analytics and advertising.' This creates both privacy and cybersecurity
risks for children and their families: in February 2026, for example, US
Senators Marsha Blackburn and Richard Blumenthal identified a public
website where any visitor could download thousands of pieces of sensi-
tive data regarding Miko's interactions with specific children.'
Finally, chatbot toys incorporate the same gamification mechanics as
many other commercial digital products, including purchases of virtual
currencies and rewards, paywalled content, and a subscription model that
limits free access to this "always available" companion to, i.e., 30
hours a month. While fee-based and paywalled models have become a
commonplace feature of digital products and services, the fact that
chatbot toys are being commercialized in this manner is particularly
disturbing, given that the product is designed to create a level of
emotional dependency amongst users that makes it extremely difficult for
a child to cease usage when they've run up against an arbitrary hours
limit.
Given all of these various risks, the Federal Trade Commission (FTC)
opened a formal inquiry into seven major chatbot developers, several of
whom have integrated their AI companions into children's toys, in
September 2025, seeking information on how the tech companies evaluate
the safety of their models, monetize user engagement, measure, test, and
mitigate negative impacts to children, both pre- and post-deployment,
use or share personal data obtained through conversations, and comply
with COPPA's prohibition on nonconsensual data collection for children
under the age of 13.6 Meanwhile, Mattel, the manufacturer of Barbie, Hot
Wheels, Fisher-Price, American Girl, Polly Pocket, and other products,
announced a partnership with OpenAI in July 2025 to "bring the magic of
AI to age-appropriate play experiences," despite OpenAI also acknowledg-
ing on its own webpage that ChatGPT is not meant for children under 13,
that all interactions with the LLM for users in this age range should be
supervised by an adult, and that the model may produce outputs that are
"not appropriate for all audiences or all ages."' 8
This bill places a temporary moratorium on the sale of chatbot toys in
New York for five years. Once the prohibition expires, lawmakers will be
able to make a decision about whether to continue the moratorium or
allow the sale and distribution of these products to children, which may
include some measure of regulation to account for their various social
isolation, dependency, and data privacy externalities outlined above.
This bill takes the logical step of putting a temporary pause on the
sale of this deeply personal yet virtually unregulated childrens product
before allowing tech developers unfettered access to our state's young-
est and most vulnerable users.
PRIOR LEGISLATIVE HISTORY:
None
FISCAL IMPLICATIONS:
TBD
EFFECTIVE DATE:
This act shall take effect on the ninetieth day after it shall have
become a law and shall expire and be deemed repealed five years from
such effective date.
1 Use Case Review: AI Toys." Common Sense Media, 2026,
www.commonsensemedia.org/ai-ratings/ai-toys section-0. Accessed 21 Feb.
2026.
2 Robb, Michael B., and Supreet Mann. Al in the Toy Box: How Parents
View AI-Enabled Toys for Young Children. Common Sense Media, 2026,
www.commonsensemedia.org/sites/default/ files/research/report/
2026-ai-inthe-box-white-paper_final-for-web-release.pdf.
3 Miko, and Miko 2. "Miko 2 - Personal AI Robot for Kids / Miko Advanced
STEM Toys." Miko 2, miko.ai/.
4 Cross, R.J., and Rory Erlich. AI Comes to Playtime: Artificial
Companions, Real Risks. Dec. 2025.
5 Collier, Kevin. "AI Toy Company Miko Adds AI Off-Switch after Poli-
tical Pressure." NBC News, 17 Feb. 2026, www.nbcnews.com/tech/ security/
ai-toy-company-miko-adds-ai- switch-political-pressure-rcna259401
www.nbcnews.com/tech/security/ ai-toy-companymiko-adds-ai-switch- poli-
tical-pressure-rcna259401. Accessed 21 Feb. 2026
6 "FTC Launches Inquiry into AI Chatbots Acting as Companions." Federal
Trade Commission, 11 Sept. 2025, www.ftc.gov/news-events/news/
press-releases/2025/09/ ftc-launches-inquiry-ai- chatbots-acting-compan-
ions?utm_source.govdelivery.
7 "Mattel and OpenAl Announce Strategic Collaboration." Mattel.com, 12
June 2025, corporate.mattel.com/ news/mattel-and-openai-announce-
strategic-collaboration.
8 OpenAI\. "Is ChatGPT Safe for All Ages? / OpenAl Help Center."
Help.openai.com, Jan. 2026, help.openai.com/en/articles/
8313401-is-chatgpt-safe-for-all-ages.
