San Jose Mayor Embraces ChatGPT, Pushes AI Integration for 1,000 City Workers

SAN JOSE, Calif. — July 17, 2025 — San Jose Mayor Matt Mahan is spearheading an ambitious push to integrate artificial intelligence into city operations, starting with the use of ChatGPT to help draft his speeches—and expanding to nearly 1,000 city employees by next year.

Already using OpenAI’s chatbot to streamline everything from public remarks to grant applications, the mayor says artificial intelligence can eliminate administrative burdens and improve public services for the city’s one million residents.

“You can knock out these tasks at a similar or better level of quality in a lot less time,” Mahan said. “But you still need a human in the loop.”

From Speeches to Budget Drafts: ChatGPT in City Hall

Mahan, who regularly speaks at public events from business openings to cultural festivals, has used ChatGPT to draft background memos and talking points. The tool has also helped shape San Jose’s $5.6 billion city budget.

Rather than shy away from public scrutiny, the mayor is open about using AI to improve efficiency—and wants other city departments to follow his lead.

By 2026, San Jose plans to train about 15% of its 7,000 city workers to use AI tools for tasks such as responding to pothole complaints, coordinating bus routes, and analyzing surveillance footage in criminal investigations.

AI in Action: Grant Writing Made Easier

One of the city’s earliest adopters, Andrea Arjona Amador from the transportation department, used ChatGPT to secure a $12 million grant for electric vehicle charging infrastructure.

She created a customized AI agent to track deadlines, organize correspondence, and help draft a 20-page proposal—saving late nights and weekends in the process.

Arjona Amador also built another tool to edit the tone and clarity of professional documents. A multilingual speaker, she said the AI assistance has significantly improved her workflow.

San Jose has so far invested over $35,000 on 89 ChatGPT licenses for city staff, at roughly $400 per user.

Bay Area Cities Leading AI Adoption

San Jose’s AI initiative mirrors similar efforts across the Bay Area. San Francisco Mayor Daniel Lurie this week announced a plan to equip nearly 30,000 city workers, including nurses and social workers, with access to Microsoft’s Copilot chatbot, also built on OpenAI technology.

San Francisco’s program includes safeguards for privacy, bias, and appropriate use. San Jose has adopted similar standards, and so far, reports no major issues—though officials remain cautious of “hallucinations,” a known problem where AI tools generate false or misleading information.

“You can’t just press a couple of buttons and trust the output,” Mahan said. “You still have to verify, use logic, and ask questions.”

Cost, Risk, and the AI Agent Challenge

Outside Silicon Valley, some cities are moving more slowly. In Stockton, IT director Jamil Niazi explored the use of AI “agents” for practical city services—like park reservations or checking pool crowd levels—but the city ultimately paused implementation due to cost concerns.

Even as tech companies tout “agentic AI” as the next frontier, adoption comes with challenges. Market research firm Gartner predicts that over 40% of AI agent projects will be scrapped by 2027 due to rising costs, unclear value, or risk concerns.

Still, Mahan remains optimistic that smart deployment of AI can make city workers 20% to 50% more productive, especially in bureaucratic-heavy departments like finance, HR, and grant writing.

“There’s just an amazing amount of bureaucracy that large organizations have to have,” Mahan said. “These tools can help us cut through that more efficiently.”

Balancing Innovation and Oversight

San Jose’s approach is to test, adjust, and remain transparent—a model Mahan hopes will spread to other government agencies.

“The idea is to try things, be really transparent, look for problems, flag them, share them across different agencies, and work with vendors and internal teams to problem-solve.”

While AI adoption in government is still in its early stages, San Jose’s experiment could become a model for how public institutions embrace technology—responsibly and at scale.