New York City creates Director of AI role

New York City’s Office of Technology and Innovation (OTI) has created a new position for a Director of Artificial Intelligence and Machine Learning and is seeking the right candidate.

The role, which offers a salary of US$75,000 to US$140,000 per year, includes responsibilities from defining what artificial intelligence (AI) means for the public and city agencies and how they can use it, to developing a comprehensive governance strategy.

Ray Legendre, a spokesperson for OTI, told Cities Today: “New York City government needs a centralised strategy around AI and machine learning to ensure city agencies effectively and responsibly use these tools to bolster their work and better serve New Yorkers.”

He added: “The private sector has accomplished a lot of great work on AI and machine learning. We’re excited to advance these efforts in the public sector. New York City aims to lead this new frontier of government-driven AI innovation and leadership – and our new director will play an instrumental role in achieving this lofty goal.”

Broader role

Former Mayor Bill de Blasio created the similar-sounding position of Algorithms Management and Policy Officer (AMPO) in 2019 but it was folded back into the Office of Data Analytics when Mayor Adams took office last year and consolidated various technology-related agencies under OTI.

New York’s Chief Technology Officer Matthew Fraser, who oversees OTI, told Cities Today in a recent interview that the AMPO workload was too big for an office of one or two people.

The city also assembled an automated decision systems (ADS) taskforce in 2018, which was the first of its kind in the country. The initiative resulted in a report of recommendations in 2019 (including to create a centralised resource) but it also received some criticism related to transparency and ultimate impact.

On the new approach, Legendre said: “The director will consider issues related to AI and machine learning more broadly, not just the set of algorithmic tools that have been commonly discussed or thought about, but broader issues, newer use cases, and applications that may be relevant to city agencies.

“Because this person sits in the consolidated Office of Technology and Innovation, they will have an opportunity to think about these issues from a citywide perspective with the ability to craft policy and discuss issues that are relevant to city agencies, while also possessing a greater degree of oversight capability than we’d have seen before the consolidation of OTI.”

Strategy

The former administration published an AI strategy in 2021, which is currently archived on OTI’s site.

“The action plan we’re hoping to produce should help us better understand where comprehensive governance is needed so that we can subsequently respond to those needs,” Legendre said.

“Our plan will build upon the important groundwork around AI and government that’s already been laid, including here in New York City.”

There are two sides of the coin,” he added. “There are opportunities where we can encourage agencies to consider AI solutions that make sense for their missions but also do so in a way that is responsible, has the most appropriate guardrails, and acknowledges the risks that sometimes come along with AI solutions.“

He said a one-size-fits-all governance model wouldn’t work, noting that issues also need to be considered at an agency-specific or solution-specific level.

Definitions

Another priority is clarifying terminology.

“AI is often used as a monolithic term but it actually refers to a large collection of technologies and tech-based solutions,” said Legendre. “Added clarity here is important when we conceptualise what this diverse set of solutions looks like and where they can plug into agency operations.”

Previous audits of algorithmic tools used by city agencies include the New York Police Department’s use of facial recognition technology and gunshot detection, a ‘severe harm predictive model’ in the Administration for Children’s Services, DNA analysis for crime scene evidence, and an application that matches applicants to schools.

AI oversight is a priority for a growing number of cities as technology and potential use cases advance. A group of nine European municipalities recently launched an open-source standard to make it easier to share information about how they use algorithms.