The Binghamton-based initiative will combine academic expertise and Empire AI computing resources to independently evaluate advanced artificial intelligence systems.
The Empire Magazine | Artificial Intelligence and Research
Information checked: 11 November 2026
The New York AI safety research centre will bring together more than 80 researchers from institutions across the state to examine the safety and security of advanced artificial intelligence systems.
Governor Kathy Hochul launched the statewide initiative on 10 November 2026 through the New York Center for AI Responsibility and Research, known as NYCAIRR, based at Binghamton University.
Its work will combine independent testing, foundational research and training for future AI safety specialists.
The centre was first announced in January 2026. The November announcement develops that foundation through statewide coordination, interim leadership and an expanded independent evaluation effort.
A Research Network Across New York
The initiative will connect scientists from colleges and research institutions throughout New York State.
Participating researchers will continue working at their home institutions. Binghamton will serve as the centre’s base within this wider network.
That structure allows the initiative to draw on expertise across universities, including specialists in computer science, security and the behaviour of complex systems.
The figure of more than 80 researchers describes the planned collaborative network. It should not be interpreted as an announcement that all those researchers are relocating to a single facility.
Who Will Lead the Centre?
Greg Morrisett, dean and vice provost of Cornell Tech, and Jeannette Wing, executive vice president for research at Columbia University, will serve as interim co-directors.
Their responsibilities include guiding the scientific agenda, overseeing independent evaluations and developing collaboration across participating institutions.
The leadership arrangement brings together researchers experienced in examining how computing systems operate, where they can fail and how their reliability can be improved.
A national recruitment effort is also planned to strengthen expertise in areas including mathematical foundations, verification, complex AI systems, privacy, interpretability and biosecurity.
What the New York AI Safety Research Centre Will Study
The centre’s objectives extend from testing existing systems to developing methods for assessing more capable future models.
Binghamton identifies several connected research priorities:
| Research priority | Intended purpose |
|---|---|
| Safety and security | Investigate whether systems behave as intended and resist misuse |
| Transparency and interpretability | Improve understanding of how AI produces results |
| Repeatable science | Develop methods, datasets and practices that support reproducible research |
| Alignment | Study ways to keep system behaviour consistent with intended values and constraints |
The statewide announcement also identifies responsible scientific discovery as part of the mission.
The aim is to produce evidence that can help researchers and public officials understand both the capabilities and limitations of advanced AI.
How Empire AI Will Support the Work
The initiative will use computing resources supplied through Empire AI, New York’s public-private programme supporting research for the public good.
The state describes Empire AI as a $500 million initiative. Its shared infrastructure gives participating researchers access to computing resources for demanding projects.
For the centre, that support could make it easier to conduct repeated experiments and compare systems under defined conditions.
Computing capacity alone does not establish that a system is safe. Its value depends on the questions researchers investigate, the quality of their methods and how clearly they communicate the limits of their findings.
Funding Behind the Centre
The January announcement identified an initial $55 million founding commitment.
This combined a $30 million philanthropic gift from donors led by Tom Secunda, a Binghamton alumnus and Bloomberg LP co-founder, with a $25 million research capital investment.
Binghamton’s published information distinguishes operational support for people and research from capital support for facilities and equipment.
The centre’s founding funding and its access to Empire AI’s wider computing programme are separate parts of the initiative’s support.
This combination is intended to provide the expertise, research environment and infrastructure needed for sustained academic work.
Why Independent AI Testing Matters
Independent evaluation can help public institutions assess claims made about powerful AI systems.
For example, a model may perform well on a benchmark while behaving differently when instructions change, information is incomplete or several tools interact. Research can help establish where performance remains reliable and where additional controls are needed.
The potential public benefit is better evidence for decisions about deployment.
Useful findings should explain what was tested, under which conditions, what happened and what remains uncertain. That makes the work easier for other researchers to scrutinise and build upon.
The centre’s emphasis on repeatable science is therefore relevant to both safety evaluation and the use of AI in scientific discovery.
Research and New York’s Wider AI Oversight Effort
The launch coincided with the opening of a state registration portal for large frontier AI developers under New York’s RAISE Act.
The research initiative is intended to provide objective evidence that can inform public policy and accountability.
Its announced scientific role includes investigating system behaviour, developing evaluation methods and identifying safety and security concerns.
As the programme develops, published research methods, findings and collaborations will help demonstrate how effectively it contributes to those goals.
Opportunities for Students and Researchers
Binghamton says the centre will support opportunities for faculty, postdoctoral researchers and graduate students, alongside undergraduate participation.
This training role could help develop expertise in evaluating AI systems, rather than focusing only on building them.
For participating institutions, the network also offers a way to connect specialised research with shared technical questions and computing resources.
Explore more global business, technology and innovation coverage in The Empire Magazine’s article archive.
Read our related coverage: us-consumer-prices-october-2026-inflation/
Follow The Empire Magazine:
Facebook | Instagram
The Empire Magazine | Crown For Global Insights
