New York is launching a state-backed AI safety research center at Binghamton University to independently test AI systems, develop safeguards and train researchers in responsible AI.

The center’s announced funding includes a $30 million philanthropic gift led by Tom Secunda and an initial $25 million investment from SUNY, with New York State covering operating costs.

Greg Morrisett and Jeannette Wing will serve as interim co-directors, setting the research agenda and building a network of researchers who will remain at their home institutions.

Its research will focus on preventing harm, anticipating failures and attacks, and improving how AI risks are detected and addressed, including risks posed by autonomous agents.

At the state’s request, the center can provide technical evidence for policymaking and work with the Office of Digital Innovation, Governance, Integrity, and Trust (DIGIT) to review developers’ published safety frameworks and recommend improvements.

Morrisett said the center’s dedicated focus on safety and security can complement companies’ work and provide independent research amid pressure to build more powerful technology.

Hochul argued that competitive pressure can give companies financial incentives to move too quickly, making independent oversight important for protecting children and New York’s economy.

Cristin Frodella of Common Sense Media’s New York parent advisory board warned that risks range from children relying on chatbots to less predictable harms that developers themselves may not anticipate.

A major constraint is that researchers can assess only information companies make public, leaving them without access to private models under development or many unreleased systems unless companies cooperate or regulators compel disclosure.

The center will conduct research but has no automatic enforcement authority, and New York has not specified whether its findings will affect licensing, procurement, enforcement or disclosure requirements.

Large frontier AI developers must register through the RAISE Portal by Jan. 1, 2027, when amended provisions of the RAISE Act are scheduled to take effect.

The Act defines large developers as those with annual gross revenue above $500 million, including affiliate revenue, and allows civil penalties of up to $1 million for a first specified violation and $3 million for subsequent violations.