Who will win the battle between robots and public-employee unions?
By Rafael Perez | February 20, 2026
For decades, artificial intelligence has been heralded for its potential to revolutionize the labor market and the creation of goods and services. Recent advancements in large language model (LLMs) performance by developers such as OpenAI and Anthropic have fueled optimism that AI will eventually eliminate the need for human workers and bring about a new age of efficiency and bounty.
Commerce is not the only domain that stands to gain from a forthcoming AI revolution — municipal functions also stand to improve.
According to many computer scientists and philosophers of artificial intelligence, AI will eventually supplant most labor currently performed by humans. Predicting the exact trajectory of AI progress is difficult, but despite persistent problems and limitations, these do not appear to be intractable problems by nature.
Even assuming only modest gains in future AI sophistication, a fairly conservative prediction is that AI will be capable of assisting in many of the routine processes happening in our cities even if they do not completely replace human labor.
Current generative LLMs are suited for language processing-based tasks such as text analysis and summary, pulling data from large sets of documents, generating content from prompts and research — all functions that will allow AI to decrease the demand for human labor.
Discriminative AI continues to make gains in concept demarcation and object classification within chaotic contexts (like road navigation). These are human skills, such as being able to discriminate between types of objects, that were once challenging barriers for AI, but that are now on-track for automation.
The benefits of this progress have already begun to be seen in cities. Researchers at Carnegie Mellon University have implemented AI-controlled traffic lights in cities including Pittsburgh, reportedly reducing travel times by up to 25% (within the areas of the pilot program) by having AI manage the flow of traffic. Cities in California including Los Angeles, San Diego and Irvine have been testing similar systems.
Likewise, Hawthorne and Bakersfield have implemented their own chatbots to assist residents in accessing services while San Francisco has made Microsoft’s Copilot available to city employees to assist in their work — chatbots replace human phone operators and AI assistance allows tasks to be performed by smaller workforces. According to San Jose’s mayor, Matt Mahan, city staff have cleared backlogs and seen their productivity increase up to 20% since the city implemented an AI training program.
AI will increasingly be used to handle resident service requests, transportation administration, utility grid management, resource and benefits allocation, planning and permitting, and will continue to expand AI’s general role in governance.
Just as AI has been observed to shorten travel times by managing traffic, there will be similar improvements in the efficiency and delivery of municipal services. Residents stand to enjoy more efficient public transportation and streamlined bureaucratic processes that cut back on what can be months of review.
As such, we can expect that AI will replace a significant portion of the municipal workforce, particularly in California where new technology is often first implemented because of its position as the premier center of innovation.
A strong labor union presence in the state also threatens to delay the possible benefits of municipal service automation for residents.
Last year, Gov. Gavin Newsom vetoed Senate Bill 1220, which was a California Labor Federation-sponsored bill aimed at preventing call centers for state and local agencies from “using or contracting for the use of AI or ADS [automated decision systems] that eliminate or automate the core job function of any worker.” Far from having the interests of the wider public in mind, bills like these are motivated entirely by unions protecting obsolete workers irrespective of the possible benefits residents may receive, such as shorter wait times and quicker access to services.
In October, Newsom vetoed Senate Bill 7 (also sponsored by the California Labor Federation), which, among other things, banned all employers, including state and local governments, from relying exclusively on ADS for “employment-related decisions.” “Employment-related decisions” and “ADS” were defined so broadly that SB 7 would have required humans to oversee any AI-derived decisions that could impact an employee in any way — virtually crippling possible sources of productivity and efficiency gains. For hiring, firing and disciplinary decisions, an entirely separate human-conducted investigation was required to corroborate AI decisions, making the use of AI entirely superfluous.
We can expect that labor unions will continue to oppose future AI implementations that threaten the municipal workforce, even if such implementation promises to improve services for residents.
Leadership in both Sacramento and in individual California cities must embrace the position of privilege they enjoy as the cradle of artificial intelligence by continuing to reject legislation that sacrifices automation in favor of protecting government workers. As these new technologies mature, it is officials’ duty to taxpayers to take every reasonable opportunity to improve their lives.
Municipal services are not taxpayer-funded jobs programs and cities have no duty to protect government employees from being made obsolete. The government’s overriding duty is to provide the services that taxpayers fund as efficiently as possible.
Rafael Perez is a columnist for the Southern California News Group. He is a doctoral candidate in philosophy at the University of Rochester.
