New York RAISE Act — Risk Assessment

Yes. New York RAISE Act addresses Risk Assessment through 1 provision.

Does New York RAISE Act require Risk Assessment?

New York • enacted

Yes — 1 provision

This regulation imposes 5 specific requirements for Risk Assessment across 1 provision:

Internal governance — Large frontier developer framework must describe internal governance practices ensuring implementation (§ 1421(1)(i)); the reviewed amended Article 44-B does not require a designated senior AI safety officer
Safety framework publication — Large frontier developers must write, implement, comply with and conspicuously publish a frontier AI framework addressing § 1421(1)(a)-(j)
Internal-use risk summaries — Large frontier developers must confidentially transmit internal-use catastrophic-risk assessment summaries to the DFS Office every three months or on another reasonable requested schedule agreed by the Office (§ 1422(2))
Annual update — Large frontier developers must review and update as appropriate at least annually; publish material modifications and justification within 30 days (§ 1421(2))
Assessment and deployment transparency — Large frontier framework must describe assessments, mitigations and review in deployment/extensive-internal-use decisions (§ 1421(1)(b)-(e)); all frontier developers publish a transparency report before or concurrently with deployment, with assessment/results/third-party-evaluator summaries for large developers (§ 1421(3))

enacted

Scope:: Large frontier developers for framework/assessment duties; all frontier developers for deployment transparency reports. Frontier models exceed 10^26 training operations including subsequent fine-tuning/reinforcement/material modifications; large developers exceed $500M prior-year gross revenue with affiliates (§ 1420(8)-(10)). Models must be developed, deployed or operating wholly or partly in New York (§ 1425); § 1426 academic-research and Empire AI/institute exceptions apply

Requirement	Details
Internal governance	Large frontier developer framework must describe internal governance practices ensuring implementation (§ 1421(1)(i)); the reviewed amended Article 44-B does not require a designated senior AI safety officer
Safety framework publication	Large frontier developers must write, implement, comply with and conspicuously publish a frontier AI framework addressing § 1421(1)(a)-(j)
Internal-use risk summaries	Large frontier developers must confidentially transmit internal-use catastrophic-risk assessment summaries to the DFS Office every three months or on another reasonable requested schedule agreed by the Office (§ 1422(2))
Annual update	Large frontier developers must review and update as appropriate at least annually; publish material modifications and justification within 30 days (§ 1421(2))
Assessment and deployment transparency	Large frontier framework must describe assessments, mitigations and review in deployment/extensive-internal-use decisions (§ 1421(1)(b)-(e)); all frontier developers publish a transparency report before or concurrently with deployment, with assessment/results/third-party-evaluator summaries for large developers (§ 1421(3))

Violation	Fine
Specified large-frontier-developer violations (§ 1427(1))	AG may seek up to $1M first / $3M subsequent, based on severity; no private action under § 1427(2)

Sources: S8828 enacted amendment text, Chapter 96 Official amendment action/status snapshot Codified Article 44-B definitions Codified §1420 revision snapshot Governor's Announcement