Law / United States / 
New York
Artificial Intelligence Training Data Transparency Act
N.Y. Gen. Bus. Law art. 44-C (proposed §§ 1430-1432), A6578-B / S6955
Finalized, with one agreed text, dated , as of .
An AI training data rule binding private bodies.
- How the hook was established
- express
- What makes it apply
- Market targeting
- Audit expectation
- none
Where it has got to
The text described here is Assembly print A6578-B, the text both houses passed. That print is A6578B, published .
Locally, this stage is passed both houses, not yet delivered to the Governor.
The stage above is recorded at www.nysenate.gov.
The Assembly passed A6578-B on and the Senate substituted it for its companion S6955-A and passed it on . In New York a bill passed by both houses reaches the Governor only when it is delivered, which can be months later; no delivery, signature or veto is recorded on the docket.
What it requires
- Publicly post documentation describing the datasets used to train a generative AI model or service made available to New York users, including their sources, whether they contain copyrighted, personal or aggregate consumer information, the collection period, and whether synthetic data generation was used.
These lines are LexLint's reading of the law's text.
In the frameworks
- NIST AI RMF
- MAP 2.3Scientific integrity and TEVV considerations are identified and documented, including those related to experimental design, data collection and selection (e.g., availability, representativeness, suitability), system trustworthiness, and construct validation.
Our reading of where this law's duties sit in each framework. It never says that running a control meets the law. How we read a framework against the law
What it makes you log
Who may demand the log
Public
What the log must hold
Data lineage
Logging duty
Section 1432(1) would require a developer to post on its own website documentation of the data used to train a generative AI model or service, before the model or a substantial modification is made publicly available to New Yorkers. It never uses the words log, record, event recording or audit trail; it requires posted documentation of the training data, which cannot be written without the developer having tracked its datasets, and the posting persists on the website rather than being a one-time disclosure at the point of output.
- Kind
- Implicit
- As of
- Provision
- Section 1432(1)
- Trigger
- content_generation
Why the legislator wanted it
Sponsor's memorandum in support of A6578-B; it also says the bill is meant to promote transparency and accountability in the development and deployment of generative AI.
- As of
- Quote
- This requirement will help identify any biases, inaccuracies, or ethical concerns in the training data that may impact the performance and fairness of Ai systems
- Source link
- https://nyassembly.gov/leg/?default_fld=&bn=A06578&term=2025&Summary=Y&Memo=Y
- Source kind
- Explanatory memorandum
- Purpose
- bias_and_performance_monitoring
- accountability_demonstration
What this law does
As passed by both houses, it would require a developer of a generative AI model or service made available to New York users to publicly post documentation describing the training data used, including its sources, whether it includes copyrighted, personal or aggregate consumer information, the collection time period, and whether synthetic data generation was used, with exemptions for aviation-safety and federal defense systems. Passed the Assembly on and the Senate on .
When LexLint raises it
When your app profile says your app trains models or generates content with AI.
Back to the example · Lint your app
Page updated .