Skip to content

Session Expired

Your session has expired. Please sign in again to continue where you left off.

Sign In Again

NY S8828: Relates to transparency and safety requirements for developers of artificial intelligence models

Enacted · set to take effect Jan 1, 2027 New York

Last verified Oct 2, 2026 · official bill text

S8828 · where it stands

4 of 5 steps complete
  1. Introduced Complete
  2. Passed first chamber Complete
  3. Passed both chambers Complete
  4. Enacted Complete
  5. Effective Jan 1, 2027
What do these statuses mean?
Introduced
Filed in the legislature; may be waiting for a committee hearing.
Passed a chamber
Approved by one or both chambers; awaiting further action.
Enacted
Signed into law, or became law without a signature or over a veto.
Effective date
The date the law says it takes effect. Once it has passed, the page shows "Effective" and the date. A court order can still block a law; if we know of one, the summary says so.
Dead / vetoed
Vetoed, failed, or the session ended without action.

In one paragraph

New York's S 8828 (Chapter 96 of 2026) is the chapter amendment to the RAISE Act (Chapter 699 of 2025). It repealed and replaced General Business Law Article 44-B and set the effective date at January 1, 2027. Developers of frontier models (over 10^26 training operations) must publish transparency reports when deploying and report critical safety incidents to a Department of Financial Services office within 72 hours. Those with over $500 million in revenue must also publish and follow a frontier AI framework, report internal-use risk assessments, file disclosure statements and pay assessments, and face Attorney General penalties up to $1 million, then $3 million. Signed March 27, 2026.

Plain-English summary, written with AI assistance. The official text controls.

What it means for business

From January 1, 2027, frontier model developers (over 10^26 training operations) whose models are developed, deployed or operating in New York must publish transparency reports and report critical safety incidents to DFS; those with over $500 million in revenue also need a frontier AI framework.

Key provisions

  1. Repeals Article 44-B as added by Chapter 699 of 2025 and enacts a new Article 44-B (§§ 1420-1429); amends Chapter 699's effective-date section from the 90th day after signing to January 1, 2027
  2. Compared with the signed text: coverage turns on a 10^26-operation training threshold, with the heaviest duties for developers with over $500 million in annual revenue (not $100 million in compute spending); reports go to a DFS office (not the Attorney General and DHSES); top penalties fall from $10 million / $30 million to $1 million / $3 million; the ban on deploying a model with an unreasonable risk of critical harm is not carried over
  3. Large frontier developers must write, implement, follow and publish a frontier AI framework covering catastrophic-risk thresholds, mitigations, third-party assessment, model-weight cybersecurity, incident response and internal use; review it yearly and publish material changes with a justification within 30 days
  4. Every frontier developer must publish a transparency report before or when deploying a new or substantially modified frontier model; large frontier developers add summaries of catastrophic-risk assessments, results and third-party evaluator involvement; model or system cards count; redactions must be explained and unredacted copies kept five years
  5. Critical safety incidents (unauthorized access to, modification of or exfiltration of model weights, or loss of control, causing death or bodily injury; harm from a materialized catastrophic risk; or a model deceiving its developer to subvert its controls, outside an evaluation, in a way showing materially increased catastrophic risk) must be reported to the office within 72 hours; one posing an imminent risk of death or serious physical injury must be disclosed within 24 hours to an appropriate authority, such as law enforcement; reports are exempt from FOIL
  6. Large frontier developers may not develop, deploy or operate a frontier model in New York without a current disclosure statement (ownership, offices, contacts; renewed every two years, on a transfer of ownership or on a material change) and paying a pro rata share of the office's costs; after notice and hearing, the office may impose $1,000 a day for not filing or not correcting false information, plus assessments owed
  7. The Attorney General may seek civil penalties up to $1 million for a first violation and $3 million for each later one when a large frontier developer fails to publish or transmit a required document, makes a materially false or misleading statement, fails to report an incident or fails to follow its own framework; no private right of action; academic research by accredited New York colleges and the Empire AI consortium are exempt; the office may adopt rules and must report yearly to the Governor and legislative leaders beginning January 1, 2028
  8. Signed March 27, 2026 (Chapter 96 of 2026); the effective-date change took effect immediately; the new Article 44-B takes effect January 1, 2027

Latest legislative action

Mar 27, 2026. Signed Chapter 96 of 2026

Sponsors

NameRole
D Primary

Topics and industries

Affected industries

This page is general information, not legal advice. Summaries are AI-assisted and may contain errors; read the statute for the controlling text and talk to a lawyer about your situation. Full disclaimer · Report an error