State Artificial Intelligence Policy
Introduction
States and municipalities are increasingly taking interest in Artificial Intelligence and filling the gaps left by federal inaction on algorithmic transparency. States and cities have taken different routes, from notification and task forces to minimum privacy standards. Some of the recent efforts are highlighted below.
State and Local Laws
Idaho Law
In March 2019, Idaho enacted a law that requires “all documents, data, records, and information used by the builder to build or validate the pretrial risk assessment tool and ongoing documents, data, records, and written policies outlining the usage and validation of the pretrial risk assessment tool” to be publicly available; allowing a party in a criminal case to review the calculations and data underlying their risk score; and precluding trade secret or other intellectual property defenses in discovery requests regarding the development and testing of the tool. This is an exemplar for states committed to using algorithms in pre-trial sentencing while retaining the notions of fairness and due process.
AI Task Forces and Commissions
- EPIC Model Commission Bill: Provides states interested in passing a law similar to the following laws a bill that prioritizes transparency. This bill has two main phases — firstly, to catalogue and publish how the state uses AI. Secondly, to recommend specific regulation about how the state should use it moving forward.
- New York City (2017): The New York City Council created a task force to study how it uses AI and to provide recommendations on specific prompts. In November 2019, the council released their report. In conjunction with this released report, Mayor De Blasio announced an Executive Order creating an “Algorithms Management and Policy Officer.” An unofficial “shadow report” of the Task Force was also released.
- Vermont (2018): The Vermont Legislature created an AI Task Force to explore areas of responsible growth of the state’s technology markets, the use of AI by their government, and appropriate regulation in the field. The task force published an update report in February 2019.
- Alabama (2019) : Alabama created an AI Commission that has a broad mandate to study “all aspects” of AI and associated technologies and the associated challenges and opportunities.
- New York State (2019): New York State created a commission to begin in 2020 that will study with a broad mandate the sufficiency of current law to deal with AI as well as the effects of AI on employment and public safety.
- EPIC Testimony on a proposed AI Commission that was being considered by the Massachusetts Legislature in October 2019.
California Bot Disclosure
California recently passed a law requiring companies to disclose when they use AI “bots” to communicate with individuals online. The law, which takes effect in 2019, follows a demonstration by Google of AI calling software, which mimics a human voice to place reservations and complete other tasks.
Washington State Proposed Notice and Legal Protections
Washington state is considering legislation that would require government agencies to conduct “algorithmic accountability” reports before adopting any automatic decision-making program. Any agency that uses automatic decision making must notify any individuals whose information is used in the system, permit audits of the data, ensure that legal rights are not impacted by the AI, and be able to explain the basis for any decision made by the program.
Resources
- EPIC Testimony on Massachusetts HB2701/SB1876, An Act establishing a commission on automated decision-making, artificial intelligence, transparency, fairness, and individual rights (Oct. 2019)
- The Public Voice, Universal Guidelines for Artificial Intelligence
- NY Times, How Do You Govern Machines That Can Learn? Policymakers Are Trying to Figure That Out (Jan 20, 2019)
- OECD, Going Digital: Artificial Intelligence
- Artificial Intelligence: The Future is Here, NCSL 2018 Legislative Summit
- Columbia Univ., ArtificialIntelligence and the Implications for Governance and Public Policy
- Brookings, Workand Social Policy in the Age of Artificial Intelligence (2017)
- GAO: ArtificialIntelligence: Emerging Opportunities, Challenges, and Implications for Policyand Research (June 26, 2018)