REMARKS
Council Member Julie Menin discusses AI governance and introduces Intro 926
0:04:11
·
156 sec
Council Member Julie Menin addresses the critical need for AI governance in New York City, highlighting the lack of an effective framework and introducing Intro 926. This legislation aims to establish responsible use practices for AI tools used by city agencies.
- Menin cites a state controller's report finding incomplete AI governance approaches in key city agencies.
- She discusses an example of an unreported AI tool used by the Department of Education.
- The proposed legislation (Intro 926) would require the creation of responsible use practices for AI tools, including fairness, transparency, and accountability measures.
Julie Menin
0:04:11
Today's hearing is critical given how technology can outpace policy.
0:04:16
Artificial intelligence in just the past year has, as we all know, become come overwhelmingly popular.
0:04:22
Earlier this year, the Pew Research Center found that nearly 1 quarter of all adults have used AI chatbot, chat GPT, and I'm one of them, I must say, Governments have also increasingly used AI since these tools can increase efficiency of tasks.
0:04:38
Unfortunately, the state controller's office found in a report that New York City, and I'll read a quote, does not have an effective AI governance framework.
0:04:47
The report additionally reviewed 4 agencies, NYPD, ACS, DOE, and DOE, and found there are incomplete approaches to AI governance.
0:04:56
For example, the Department of Education uses an AI tool called teach to 1 360 to identify students problem areas and strengths.
0:05:06
Specifically, it's reported that that tool then creates a personalized daily lesson for students.
0:05:13
This tool, however, was not reported to the algorithm's management and policy office under the mayor's office of operations as established an executive order 50 of 2019.
0:05:24
That is one of the many reasons why introduced intro 926, which would require the creation of appropriate and responsible use practices for artificial intelligence tools that are utilized by city agencies.
0:05:38
Specifically, the legislation would allow an agency such as the Department of Information Technology And Telecommunications to create rules that establish minimum practices for agencies to follow and follow compliance with AI tools.
0:05:53
This could include ensuring fairness, transparency, and accountability, and AI decision making processes such as testing against bias or identifying and mitigating risks such as data protection.
0:06:06
These practices will be reviewed annually and updated along with a report on compliance every 2 years.
0:06:13
There are many more cases of AI tools that are creating unintended bias.
0:06:17
If our city is using AI tools, then we absolutely must have proper standards and clear oversight.
0:06:24
I thank all who've joined me on this legislation, including our chair, Gutierrez.
0:06:29
Thank you so much.
0:06:30
As the city needs robust guidelines to ensure that we are properly monitoring and reviewing AI tools.
0:06:36
I wanna thank the chair for allowing me to speak, and I also wanna thank the builder after Connor Mealy, and for my team, Jonathan Shutt, Brandon Jordan, and Mercedes Anderson.
0:06:44
Thank you so much, and I look forward to today's hearing.
