Massachusetts Senate weighs AI safeguards bill

The Massachusetts state Senate debated legislation that would establish guardrails for AI as part of a larger $325 million economic development bill. The package also includes funding for sectors such as defense, age limits on e-bikes and automatic two-family zoning requirements.

The bill authorizes $75 million to “support the development and application of AI,” while requiring large AI developers to identify potential catastrophic risks from their systems and create safety frameworks to mitigate them. The risks include scenarios such as the death of over 50 people or damages of more than $1 billion from cyberattacks or assistance in creating a nuclear weapon.

Catastrophic risk is defined as a “foreseeable and material risk” from the development, storage or use of AI that could kill at least 50 people or cause $1 billion in damage in a single incident. The definition also covers expert-level assistance involving chemical, biological, radiological or nuclear weapons, as well as murder, assault, extortion, theft and systems that might “evade the control” of their developer or user.

The attorney general would be allowed to sue developers that violate the safety requirements. A commission led by the attorney general or a designee would also study AI risks and the safety protocols created by developers.