New York’s AI Safety Bill Loses Strength Amid University Opposition
A coalition of tech firms and academic institutions recently invested between $17,000 and $25,000 in an advertising campaign opposing New York’s significant AI safety legislation, potentially reaching over two million individuals, as reported by Meta’s Ad Library. This legislation, known as the RAISE Act (Responsible AI Safety and Education Act), was signed into law by New York Governor Kathy Hochul just days ago. The law mandates that AI companies developing large models—like OpenAI, Anthropic, Meta, Google, and others—must establish safety protocols and transparency measures for reporting major safety incidents to the attorney general. However, the version signed by Hochul differs from the one that passed the New York State Senate and Assembly in June, as it was revised to be more favorable to tech companies. A group of over 150 parents had urged the governor to approve the bill without alterations, while the coalition of tech firms and academic institutions, known as the AI Alliance, worked to soften its impact.
The AI Alliance, which spearheaded the ad campaign, includes major players like Meta, IBM, Intel, Oracle, Snowflake, Uber, AMD, Databricks, and Hugging Face. This isn’t surprising given their vested interests. In June, the group expressed its “deep concern” about the bill in a letter to New York lawmakers, labeling it “unworkable.” The alliance also comprises various colleges and universities worldwide, such as New York University, Cornell University, Dartmouth College, Carnegie Mellon University, Northeastern University, Louisiana State University, and the University of Notre Dame, along with engineering schools like Penn Engineering and Yale Engineering.
The ads, which launched on November 23, carried the message, “The RAISE Act will stifle job growth,” arguing that the legislation would hinder New York’s tech ecosystem, which supports 400,000 high-tech jobs and significant investments. They advocated for a future where AI development is open, trustworthy, and beneficial for the Empire State.
When inquiries were made about their unintentional involvement in the ad campaign against the AI safety legislation, none of the academic institutions responded, except for Northeastern, which did not provide a comment by the time of publication. In recent years, OpenAI and its competitors have increasingly engaged with academic institutions, forming research partnerships and offering technology to students at no cost.
While many academic institutions in the AI Alliance do not have direct partnerships with AI companies, some do. For example, Northeastern’s collaboration with Anthropic this year granted access to Claude for 50,000 students, faculty, and staff across 13 campuses globally. In 2023, OpenAI supported a journalism ethics initiative at NYU, and Dartmouth announced a partnership with Anthropic earlier this month. Additionally, a professor from Carnegie Mellon University serves on OpenAI’s board, and Anthropic has funded initiatives at Carnegie Mellon.
The original RAISE Act stipulated that developers could not release a frontier model if it posed an unreasonable risk of critical harm, defined as the death or serious injury of 100 or more individuals, or damages exceeding $1 billion related to the creation of a chemical, biological, radiological, or nuclear weapon. This definition also applied to AI models that operate without meaningful human intervention and would be considered criminal if performed by a human. The version signed by Hochul removed this clause, extended the deadline for reporting safety incidents, and reduced penalties, among other modifications.
The AI Alliance has previously lobbied against AI safety regulations, including the RAISE Act, California’s SB 1047, and President Biden’s AI executive order. Its mission is to unite builders and experts from various fields to collaboratively and transparently tackle the challenges posed by generative AI and ensure its benefits are widely accessible, particularly through member-driven working groups. Beyond lobbying, the group has also focused on cataloging and managing “trustworthy” datasets and establishing a ranked list of AI safety priorities.
The AI Alliance wasn’t the only entity opposing the RAISE Act through advertising. A pro-AI super PAC called Leading the Future, backed by various investors, has also targeted the bill’s cosponsor, New York State Assemblymember Alex Bores, with ad spending. However, while Leading the Future operates with a clear agenda, the AI Alliance functions as a nonprofit partnered with a trade association, aiming to develop AI in a collaborative, transparent manner with an emphasis on safety, ethics, and the greater good.
