How State Leaders Can Put People First in AI Decision-Making

Source Domain: fas.org

Here is a summary of the article using an unordered list with 6 key points:

– The article outlines a framework for state leaders to follow a people-first approach when making decisions about the use of artificial intelligence (AI). It emphasizes the need to consider fairness, accountability, evidence-based inquiry, and inclusive governance when deploying AI systems.

– Many state agencies are already using or considering using AI for consequential decisions in sensitive areas like healthcare, housing, education, etc. However, concerns have been raised about the risks of improper or biased AI use, which can cause harm to individuals and communities.

– The public has deep concerns about how AI will impact their lives, with polls indicating that a majority do not trust technology companies to properly regulate AI. There are also fears that low-income and marginalized communities are often excluded from decisions about government use of AI, and bear the brunt when AI fails.

– The article provides a four-step people-centered AI decision-making process for state leaders to follow before deploying any AI system:

1) Gather inclusive input from diverse community members
2) Conduct an AI Impact Assessment of benefits and risks
3) Use a decision-making standard based on community considerations and evidence
4) Evaluate the AI system’s performance to determine if it continues to be beneficial

– Examples of dangerous AI uses that should be prohibited include social scoring, facial surveillance, and uses in predictive policing.

– The article provides recommendations for how state leaders can operationalize a people-first AI decision-making process through executive actions, legislation, agency guidelines, or a combination of these mechanisms. It also suggests providing structures to support state agencies in implementing the process.