Massachusetts Bill S. 31

• Authority: Massachusetts
• Parent authority: State governments
• Jurisdiction: United States
• Collections: U.S. state and local documents
• Most recent activity: Proposed
• Most recent activity date: 2023-01-20

Risk subdomain coverage

Good coverage — Comprehensive governance measures or mitigations with clear procedures relating to the risk subdomain are explicitly described.

• Compromise of privacy
• Lack of transparency or interpretability

Minimal coverage — Brief mention but no specific mitigations, recommendations or governance measures relating to the risk subdomain are described.

• Unfair discrimination and misrepresentation
• AI system security vulnerabilities
• Disinformation, surveillance, and influence at scale
• Fraud, scams, and targeted manipulation
• Economic and cultural devaluation of human effort
• Lack of capability or robustness

All other risk subdomains had no coverage in this document.

Risk Coverage Summary

The document has good coverage of approximately 8-9 subdomains, with strong focus on discrimination (1.1), privacy compromise (2.1), security vulnerabilities (2.2), disinformation/surveillance (4.1), fraud/manipulation (4.3), lack of robustness (7.3), and lack of transparency (7.4). Coverage is concentrated in discrimination, privacy/security, malicious use prevention, and AI system reliability domains.

Sector coverage

Good coverage — Comprehensive governance measures or mitigations with clear procedures relating to the sector are explicitly described.

• Information

Minimal coverage — Brief mention but no specific mitigations, recommendations or governance measures relating to the sector are described.

• Scientific Research & Development Services

All other sectors had no coverage in this document.

Sectors Summary

This bill governs AI developers across all sectors by regulating companies operating large-scale generative AI models. The regulation is sector-agnostic and applies to any company meeting the technical threshold (≥1 billion parameters), regardless of industry. The primary sector directly governed is Information (AI model developers), but the requirements apply to any organization operating such models.

Legislative Status

Hard Law

Legislative Status Summary

This is a binding legislative bill with mandatory requirements, enforcement mechanisms through the attorney general, and legal penalties for non-compliance under chapter 93A.

Actors

Proposer

• Entity Types: AI_Governance_Actor
• Entities: Massachusetts Legislature
• Summary: This is a Massachusetts state bill (S.31) proposed through the legislative process, as indicated by the bill number and structure following Massachusetts General Laws format.

Target

• Entity Types: AI_Developer
• Summary: The bill explicitly targets companies that operate large-scale generative AI models with at least one billion parameters, which are AI developers/providers.

Enforcer

• Entity Types: AI_Governance_Actor
• Entities: Massachusetts Attorney General
• Summary: The Attorney General is explicitly designated as the enforcement authority with power to adopt regulations and bring enforcement actions.

Monitor

• Entity Types: AI_Governance_Actor
• Entities: Massachusetts Attorney General
• Summary: The Attorney General is responsible for maintaining a public registry of registered companies and monitoring compliance through the registration system.

AI Lifecycle Stages

Lifecycle stages covered by the documents, as defined in the OECD Framework for the Classification of AI Systems

• Plan and Design
• Collect and Process Data
• Build and Use Model
• Verify and Validate
• Deploy
• Operate and Monitor

AI Lifecycle Summary

The document covers multiple lifecycle stages with primary focus on Build and Use Model (through design and development requirements), Deploy (through registration requirements), and Operate and Monitor (through ongoing risk assessments and compliance obligations). It also addresses data collection and processing requirements.

Technical Scope

Terms relating to technical scope that are mentioned or implicitly referenced in the document

• AI Models
• Compute Threshold
• Generative AI

Technical Scope Summary

The document explicitly defines and covers large-scale generative AI models with at least one billion parameters. It specifically mentions ChatGPT as an example. It does not distinguish between frontier AI, general purpose AI, or task-specific AI, nor does it mention foundation models, predictive AI, open-weight models, or specific compute thresholds beyond the parameter count.