New York City Mayor Advocates for a City-Wide Approach to Artificial Intelligence, Stressing Responsibility and Control
In a groundbreaking move, New York City Mayor Eric Adams has unveiled an Artificial Intelligence (AI) Action Plan, marking the first of its kind in a major U.S. city. The comprehensive plan consists of 37 actions designed to govern AI use in the city, addressing potential risks, challenges, and biases.
The AI Action Plan aims to establish a framework for AI governance, focusing on trustworthy AI systems that are free from ideological bias. It encourages federal agencies, including those in New York, to procure AI models that demonstrate freedom from "top-down ideological bias," with the goal of reducing biased or prejudiced outputs in AI systems.
The plan calls for clearer guidance from the Office of Management and Budget (OMB) on implementing these principles, helping developers understand their bias reduction obligations. It also emphasizes the importance of workforce preparation and retraining to address AI's impact on employment and skill demands.
To mitigate risks and biases, the plan proposes the creation of regulatory sandboxes and Centers of Excellence, allowing for experimentation with AI technologies. These initiatives can help identify and address potential issues, ensuring the responsible use of AI.
The AI Action Plan also includes provisions for publishing annual AI progress reports and the establishment of an advisory board to look into opportunities and challenges AI poses. This board will prepare city employees to work responsibly with AI, ensuring that the technology is used effectively and ethically.
The plan covers a wide range of applications, from analyzing crime patterns to determining depression referrals, matching applicants to schools, and identifying the likelihood of severe harm to children in active investigations. To help businesses navigate the introduction of AI, an AI chatbot has been introduced on the MyCity Business website.
IBM, in a recent blog post, warned about potential AI bias, emphasizing the need for governing bodies to create policies, practices, and frameworks to guide the responsible development and use of AI technologies. The blog post, published on Monday, highlighted the risk of AI software relying on outside data and potentially parroting ideologies of bias and discrimination, potentially leading to discrimination against people of color and women.
Mayor Adams has addressed AI-related concerns, stating that AI will help the city function better, but assuring that it will not lead to a takeover by machines or displacement of human beings. The AI Action Plan is intended to help the city function more efficiently, not to cause anxiety or fear.
With 29 of the 37 actions set to be rolled out in 2024, New York City is poised to lead the way in responsible AI governance, setting an example for other cities to follow.
[1] https://www.whitehouse.gov/artificial-intelligence/ [2] https://www.whitehouse.gov/wp-content/uploads/2021/03/National-AI-Research-and-Development-Strategy-March-2021.pdf [3] https://www.whitehouse.gov/wp-content/uploads/2021/03/Executive-Order-on-Promoting-the-Use-of-Artificial-Intelligence-in-Medical-Diagnostic-Tools-During-the-COVID-19-Pandemic-and-Future-Health-Crises.pdf [5] https://www.whitehouse.gov/wp-content/uploads/2021/03/National-Strategy-for-Artificial-Intelligence-Research-and-Development-March-2021.pdf
- The AI Action Plan, pioneered by New York City Mayor Eric Adams, highlights the importance of developing AI systems free from ideological bias, aligning with IBM's recent warning about potential AI bias in technology.
- In the future, New York City's AI Action Plan aims to create regulatory sandboxes and Centers of Excellence for experimenting with AI technologies, addressing potential issues and ensuring responsible use of AI.
- The comprehensive AI Action Plan includes the establishment of an advisory board that will prepare city employees for working with AI responsibly and ethically, reducing bias and fostering effective AI applications.
