Oakland Taps AI to Modernize City Services

The global surge in artificial intelligence is no longer confined to the private sector. The city of Oakland is now cautiously exploring how these advanced tools can reshape its public services, marking its entry into what many believe is the biggest technological shift in decades.

For over a year, Oakland officials have been quietly collaborating with a local university to identify opportunities and establish guidelines for AI integration. Some city workers have already started using these tools, prompting the city to take a more structured approach.

Oakland's Cautious Approach to AI Adoption

Deputy City Administrator Joe DeVries revealed that an AI working group was formed about 18 months ago to explore the technology's potential. Recognizing the inherent risks for a public agency, the group’s first step was to develop interim security guidelines. "Clearly, there’s a lot of fear and concern as a public agency," DeVries noted. "We wanted to make sure we did it right."

Soon, the city plans to invite companies to pilot AI tools across various departments. Oakland intends to partner with these firms, offering valuable feedback on their products before they hit the market.

30 AI Use Cases on the Table

A draft document highlights 30 potential "use cases" for AI, offering a glimpse into the city's most pressing challenges. Some of the innovative ideas include:

• Chatbots to assist residents with permits, code enforcement, and pothole complaints.
• A tool to help the police watchdog scan body camera footage more efficiently.
• Software for the Finance Department to prevent "revenue leakage."
• An AI agent to match adoptable animals with potential owners at Oakland Animal Services.
• AI-enabled drones for the fire department to monitor fires.
• Technology to scan surveillance footage to identify illegal trash dumpers.
• A platform to review campaign finance reports for potential corruption red flags.
• A tool to accelerate the review of development proposals.

DeVries emphasized that for many of these applications, "obviously a human has to do the final look."

Establishing AI Rules But Facing Challenges

The city’s working group issued guidance in December 2024 on the ethical use of AI. Key rules include obtaining permission before feeding city data into an AI and citing AI usage in all public records. Staff are prohibited from using AI to access unauthorized information or inputting any personally identifiable information. The guidelines also forbid the use of any non-public city data.

However, adherence to these rules appears inconsistent. An AI-generated image of City Hall used in a press release lacked the required citation. Furthermore, chat logs with ChatGPT showed that staff had entered personally identifiable information, which was later redacted by the city before release.

Oakland has also adopted an AI "equity statement," committing to using technology that reduces racial biases—a known concern with many AI tools—and prioritizes privacy, transparency, and community engagement.

A Strategic Partnership for Innovation

In a workshop held at Northeastern University’s Mills College campus, staff from about 25 city departments brainstormed 60 potential AI uses, which were later narrowed down to 30 for a formal "request for information." The city hopes to partner with firms for 16-week pilot projects to test these ideas.

Northeastern is also providing graduate students to help city departments use tools like Microsoft 365 Copilot. This collaboration aims to study usage patterns and challenges to inform a broader city-wide rollout.

Mayor Barbara Lee's office is also involved, supporting initiatives like a city "tech week" and an "innovation lab" at Northeastern, funded by a $2.14 million grant from the Kapor Foundation. "The opportunity to use AI to benefit Oaklanders is there for the taking, and we are taking steps to ensure we do that in a responsible, ethical, and equitable way," said spokesperson Justin Phillips.

The Human Element How AI Affects City Workers

The potential for AI to displace human workers is a significant concern. DeVries stated that the city aims to support its workforce through education and training in AI, hoping it will "allow them to do more with less, which is a reality we’re facing in these tough budget times."

A survey of city employees revealed mixed feelings. Out of 85 respondents, 55 expressed concerns about data leaks and job displacement. Across the U.S., about one in five workers now uses AI in their job. Other worries included the environmental impact of data centers, which require enormous amounts of electricity, and the risk of AI replicating existing societal biases.

Conversely, 30 workers saw potential benefits, such as customer service chatbots, quick translation of documents, and streamlining the permitting process. Even these optimists urged caution. As one employee wrote, "If the city is to use AI, I think it needs to be for simple, clerical processes. I don’t believe we should be using it for content generation or policy purposes at this phase."