New York's legislature doesn't usually end its session with consensus on anything. Three AI bills passing on votes of 137-0, 60-0, and near-unanimous margins is unusual enough that employers who have been watching the state's AI policy activity need to pay attention.
The session ended in late June 2026 with three AI bills passed and waiting for Governor Kathy Hochul. None had been delivered to her by mid-October. Two of the three directly affect how New York employers use AI in workforce and customer-facing contexts.
TL;DR: NY passed a kids chatbot ban (passed unanimously), an AI workforce reporting requirement (50+ employees file annual reports), and a generative AI accuracy notice requirement (up to $1,000 per violation). All three await Hochul's signature. For NY employers who are already managing NYC Local Law 144 AI hiring audit obligations, these bills would expand the state's AI compliance footprint significantly, and a signature could come with little warning.
The three bills
Bill 1: S9051 / A10379B, Kids Chatbot Safety Act
Passed: 137-0 in the Assembly, 60-0 in the Senate. Unanimous on both sides.
The bill prohibits operators of AI companion chatbots from providing certain features to minors. Specifically, companion chatbot operators must use age assurance methods to determine if a user is under 18 before enabling features that:
- Simulate an emotional relationship with the user
- Use personal data or behavioral patterns to foster continued engagement
- Mislead users into believing they are interacting with a human
- Generate content promoting self-harm, suicide, substance abuse, or sexual activity
The age assurance obligation sits with the operator, not the user's parent or school. If your company runs any consumer-facing AI product with companion, character, or agent features, you are the covered entity.
The penalty structure is significant: the NY Attorney General can bring legal action, seek restitution, and impose civil penalties up to $25,000 per violation. The AG, not a private plaintiff, is the enforcement mechanism.
Companion AI rules are piling up elsewhere too. China's measures for anthropomorphic AI interaction services, in force since July 15, 2026, bar providers from making it a service goal to replace social contact or induce dependence, and require a reminder after every two hours of continuous use. New York's bill adds the minor-protection and age assurance obligations to a growing stack of companion AI requirements.
Who needs to act now: Any company operating a consumer-facing AI product with memory, personality, or emotional engagement features in New York.
What to do before the bill is signed: Map which of your AI product features would qualify as "companion features" under the bill's definitions. Age assurance at account creation is the minimum implementation requirement. If you serve users who might be under 18, document your current approach and identify the gap.
Bill 2: AB 9581B / S8706-B, AI Workforce Impact Reporting
The bill would amend the New York Labor Law to require "covered businesses", companies doing business in New York that employ more than 50 people or are publicly traded, to file annual reports with the NY Department of Labor by March 1 each year.
The report must describe the company's AI use and its impact on the workforce during the prior calendar year, including:
- An estimate of employees displaced due to AI adoption
- An estimate of employees whose hours were reduced due to AI
- An estimate of employees hired or whose hours increased due to AI
- Previously filled positions the company chose not to refill because of AI
That last item is the most novel: companies that replaced headcount with AI, rather than laying off existing workers, would need to report those unfilled positions as an AI workforce impact.
The report is filed with the Department of Labor, not published publicly. There is no explicit penalty provision in the available bill text for failure to file, though Labor Law violations in New York carry standard administrative penalties.
This reporting obligation is a direct response to the pattern documented nationally: companies are not laying off workers in visible AI-driven reductions, they are not refilling positions. New York's legislature wants to see that data at scale.
Who needs to act now: Companies with 50+ employees doing business in New York, or any publicly traded company with New York nexus, regardless of employee count.
What to do before the bill is signed: Set up a tracking mechanism for AI-related workforce decisions now, before the first March 1 reporting deadline. The report requires estimates, not exact counts, but the data is much easier to produce if you've been tracking it throughout the year rather than reconstructing it at year-end. HR systems that log position reclassifications, role eliminations, and hiring decisions should be configured to capture AI adoption as a factor.
Bill 3: AB 3411B / S934A, Generative AI Accuracy Notices
The bill passed the Assembly on January 28, 2026 and the Senate on March 9, 2026, months before the other two (Assembly bill record). It requires the owner, licensee, or operator of a generative AI system to "clearly and conspicuously display a notice on the system's user interface that the outputs of the generative artificial intelligence system may be inaccurate."
The penalty is a civil penalty of up to $1,000 for each violation, and each user who isn't given the notice counts as a separate violation. If signed, the bill takes effect on the 90th day after it becomes law.
The bill's scope is intentionally broad. It covers any "owner, licensee, or operator" of a generative AI system, which would include companies that license an AI system from a vendor and use it to generate outputs for their own users or employees. Whether it applies to purely internal use is a scope question that Hochul's signing statement or subsequent guidance will need to address.
This bill is the New York version of a disclosure requirement that the FTC's July 2026 proposed policy is approaching from the federal angle. The FTC is saying vendors who misrepresent output accuracy may be violating Section 5. New York is saying operators who fail to disclose the possibility of inaccuracy may owe a penalty for every user who misses the notice.
Who needs to act now: Any company that uses a generative AI system to produce outputs visible to external users, customer service chatbots, AI-generated reports, AI-assisted emails, AI-generated legal or compliance analysis delivered to clients.
What to do before the bill is signed: Audit your customer-facing AI outputs and determine where a disclosure notice would need to appear. The "clear and conspicuous" standard is the same used in FTC disclosure requirements, it means visible, not buried in terms of service. A one-time terms-of-service disclosure at account creation likely does not satisfy it for ongoing outputs.
The NYC Local Law 144 overlap
New York City's Local Law 144, which has been active since 2023, requires AI bias audits for automated employment decision tools used in hiring or promotion decisions for NYC employees. It covers a narrow slice of AI use: hiring and promotion decisions using automated tools.
The three new bills expand the state-level AI compliance footprint significantly, and the scope differences matter.
| Obligation | Scope | Applies to |
|---|---|---|
| NYC LL144 bias audit | Hiring/promotion decisions, NYC employees | Employers using automated employment tools in NYC |
| NY AI workforce reporting (pending) | Any AI use affecting workforce | 50+ employee companies, publicly traded companies, NY nexus |
| NY kids chatbot prohibition (pending) | AI companion features for minors | Consumer-facing AI companion operators |
| NY generative AI notices (pending) | AI-generated outputs | Operators of generative AI systems displaying output to users |
A company with more than 50 employees that uses AI for hiring and also runs a consumer AI product would potentially face obligations under all four frameworks simultaneously.
Timing: why to prepare before a signature
We couldn't find an official date by which Governor Hochul must act, and none of the bills had been delivered to her by mid-October. A signature could still come quickly once they are, and the compliance prep window is short for several of these bills.
The workforce reporting bill requires annual data collection starting in the reporting year. If the bill is signed late in 2026, companies may need workforce impact data for the full 2026 calendar year, data collection they should have started in January, with little time to implement tracking before the first March 1 filing deadline.
The kids chatbot bill's $25,000 per violation penalty is enough that waiting for a signed bill before implementing age assurance is a real business risk. A bill that passes 137-0 and 60-0 with unanimous legislative support is a strong signal of Hochul's likely action.
The generative AI notice requirement is the most straightforward to implement, a disclosure notice on AI outputs, but its scope questions mean legal review now will prevent scrambling later.
Practical sequence for New York employers:
- Identify whether the 50-employee threshold applies to you and whether any public listing creates New York nexus for the workforce reporting obligation.
- Map which AI products or internal tools would be covered by the generative AI notice requirement.
- If you have any consumer-facing AI with companion, agent, or personality features, assign ownership of the age assurance implementation now.
- Set up the workforce impact tracking mechanism before Hochul acts, so the 2026 data is available for the first March 1 report.
- Monitor Hochul's action on each bill separately, she can sign some and veto others, and signing statements may clarify scope on contested provisions.
New York employers who have already built NYC Local Law 144 compliance into their hiring processes have the organizational muscle to absorb new AI compliance obligations faster than companies starting from scratch. The question is whether the compliance infrastructure that handles a narrow automated employment tool obligation can expand to cover workforce reporting, consumer-facing AI disclosures, and companion AI age assurance at the same time.
Correction, October 15, 2026: an earlier version called China one of "three other states" with companion AI laws and said its rules require "anti-addiction systems and emotional boundary controls"; it also gave Rhode Island's AI therapy chatbot law an effective date of June 22, 2026, which we couldn't confirm. China's measures (in force July 15, 2026) bar dependence and addiction as service goals and require two-hour usage reminders; the Rhode Island reference was removed.
Correction, October 15, 2026 (second): an earlier version said the generative AI notice bill (AB 3411B) passed in the June session, quoted its notice as "may be inaccurate, incorrect, or misleading" and gave a flat $1,000 penalty per output. It passed both houses by March 9, 2026; the bill text says outputs "may be inaccurate"; the penalty is up to $1,000 per violation, counted per user; and it would take effect 90 days after becoming law.
Correction, October 15, 2026 (third): earlier versions, including the title, said Governor Hochul has until December 31, 2026 to act on the three bills. We couldn't confirm that date from an official source: the bills hadn't been delivered to her, and the New York Senate's bill pages didn't load for us. We removed the date.
Related Reading
-
AI workforce transition plan for small teams, including New York's WARN automation checkbox
-
California SB 947 and SB 951: No Robo Bosses and Worker Displacement Act
-
Rhode Island's AI Therapy Chatbot Laws: What the June 2026 Signing Means
-
China's AI Companion Law Takes Effect July 15: What It Requires
Sources: Fisher Phillips, "New York Employers Need to Prepare for 3 New AI Laws", Transparency Coalition, "New York Lawmakers Pass Kids AI Chatbot Safety Bill and Two Transparency Acts", Workplace Privacy Report, "AI in the Empire State: Two Bills Could Create New Compliance Obligations", Multistate.ai, "New York AI Legislation 2026: Chatbots, Transparency and Pricing", Bloomberg Law, "AI Chatbots For Minors To Face New Restrictions in NY Bill".
