Governor Kathy Hochul launched a new state-run AI testing program on Tuesday, marking the first such initiative in the United States. According to Engadget, the goal is to have independent researchers verify the work of AI companies rather than accepting their self-assessments.reportedThe state will check AI companies' work using independent researchers instead of trusting their own safety claims.New York is establishing an independent evaluation framework for advanced AI systems, leveraging academic resources to validate industry safety assertions. This shift places empirical validation above corporate claims for advanced systems operating within the state.
The figures
Engadget reports that more than 80 New York scientists will contribute to the testing efforts.reported The program utilizes computing power from Empire AI, a public-private partnership valued at $500 million. The RAISE Act mandates that large AI developers register with the state by January 1, 2027.reported Companies must report critical safety incidents within 72 hours once the law takes effect.reported
Behind the figures
The New York Center for AI Responsibility and Research at Binghamton University will handle the testing workload. These researchers will use resources from Empire AI to evaluate advanced AI systems. The center aims to study system risks and provide evidence to inform public officials. This structure ensures that technical assessments are conducted by academic institutions rather than industry insiders.
What it means for your stack
Developers should anticipate that their models may be subject to independent scrutiny if they operate in New York. The RAISE Act will require the publication of safety frameworks and transparency reports.reported While the testing center can only study what companies disclose publicly, the legal requirement for registration and incident reporting creates a new compliance layer. Engineers must prepare their documentation for external review.
Building on this
Inference engineers can expect a new layer of validation for models deployed in New York. The documented capability of the Binghamton center is to test advanced systems and study their risks. This means developers must ensure their safety frameworks are robust enough for independent analysis. The interface for this interaction is the transparency report required by the RAISE Act. Limits exist because the center can only evaluate publicly disclosed models.
Why it matters
- Developers must prepare transparency reports and safety frameworks for independent review by state researchers.
- Critical safety incidents must be reported within 72 hours, creating a strict timeline for incident response documentation.
- The use of Empire AI computing resources suggests the state can handle resource-intensive model evaluations.
- Registration with the state is mandatory for large AI developers by January 1, 2027, establishing a clear compliance deadline.
Topics: AI regulation, safety testing, New York, Binghamton University, compliance
Updated
Then / now / next
AI companies were largely responsible for policing their own safety claims.
New York has launched an independent testing program to verify AI work.
Watch for the implementation of the RAISE Act registration requirements in 2027.
Story timeline
- Then AI companies were largely responsible for policing their own safety claims.
- 10 Nov 2026 New York will independently test AI models
- Now New York has launched an independent testing program to verify AI work.
- Next Watch for the implementation of the RAISE Act registration requirements in 2027.
The strongest doubt
Counterpoint from the sources
Engadget notes a potential hole in the program: it can only study what AI companies disclose publicly, leaving unreleased models outside the scope of testing.
For readers
Developers
- Prepare safety frameworks and transparency reports for independent state review.
- Ensure incident reporting processes can meet the 72-hour deadline for critical issues.
Policy
- Monitor the registration process for large AI developers under the RAISE Act.
- Review how independent testing results inform public official decisions.
Researchers
- Contribute to the testing efforts at the Binghamton University center.
- Utilize Empire AI computing resources for advanced system evaluations.
Ask the article
Local keyword search of this article; no generated answers.Did this article give you the evidence you needed?
Your choice stays on this device. No public tally.Terms in this story
Empire AI
A $500 million public-private AI partnership providing computing power for state research.
RAISE Act
New York legislation requiring large AI developers to register and report safety incidents.
Binghamton University
The institution hosting the New York Center for AI Responsibility and Research.
Questions readers ask
Who will conduct the AI testing in New York?
The New York Center for AI Responsibility and Research at Binghamton University will handle the testing.
When do large AI developers need to register with the state?
They must register by January 1, 2027, under the RAISE Act.
What happens if a critical safety incident occurs?
Companies must report critical safety incidents within 72 hours.
Connected stories
Reading receipt
- New York established the nation's first state-run AI testing program on Tuesday.
- The New York Center for AI Responsibility and Research at Binghamton University will lead the effort.
- The RAISE Act requires large developers to register with the state by January 1, 2027.