California Unions Fight AI Efficiency

By James Erwin

Government workers and the unions that represent them are always in the business of making government more expensive. It is a self-evident truth that it is in their interest to inflate costs with boosted salaries, larded pensions, and excessive paid vacation. It makes sense, then, that they would oppose efficiencies that artificial intelligence could bring to government, saving taxpayer dollars in the process.

The California Federation of Labor Unions has initiated a lobbying campaign to impose new regulations on use of AI in hiring and workplace assessment. Governor Gavin Newsom vetoed a bill that would have implemented some of these policies last October. Ostensibly, the union bosses are fighting use of AI in hiring decisions and performance reviews to protect workers from the possible errors an impersonal decision-maker could commit. In reality, they are resisting accountability for their members, most of whom are government workers paid by the people’s taxes.

The ever-growing labor conglomerate AFL-CIO has absorbed dozens of smaller unions over the years, with the American Federation of Government Employees (AFGE) one of its largest subsidiaries. Less than 10% of American workers belong to a union, but the rate for the private sector is even lower at just 5%. Unionization is largely driven by the one-third of government employees at the local, state, and federal level that belong to AFGE-affiliated unions. While the union bosses lobbying the California legislature do represent private-sector workers (albeit a tiny minority of them), the driving force behind their unions are taxpayer-funded government employees, who are the chief beneficiaries of any concessions unions win from politicians.

Prohibiting AI in hiring (and less defensibly in performance reviews) is less about preventing mistreatment of your average job-seeker and more about insulating bureaucrats from accountability. And while algorithms playing a role in hiring might sound scary in the abstract, it likely means that decisions will be far more data-driven. In fact, replacing human judgement with all its biases and prejudices, most of them interpersonal, with hard metrics will reduce nepotism, discrimination, and otherwise unfair favoritism in hiring and firing. Those who do their jobs competently have nothing to fear. That government employees feel the need to outlaw metric-driven performance reviews is rather telling.

There are several reasons why government employees should be excited about some of the innovations AI is bringing to the workplace. As we wrote two years ago, AI adoption in the executive branch could help federal agencies better organize and keep records, such as tax records for the IRS, secure. It can can simplify complex tasks, provide insights into data, build capacity, and improve workflows. Redacting documents is significantly faster with the help of AI tools. In short, AI can make their jobs easier while improving data security, response times, and constituent service.

There are plenty of government workers happy about these improvements. County clerks have been quick to develop use cases that reduce burdens on understaffed offices. This revealed preference runs counter to the narrative pushed by the unions representing them, which seems to be that all government workers should be above accountability at all times.

What they likely fear is automation of their jobs. While fears of AI taking knowledge workers’ jobs have so far proven a bit overblown, the two exceptions seem to be in tech companies (chatbots are really good at coding and sparked a round of layoffs in the last three years) and the federal government, which has shed over 300,000 full-time positions since 2024. Of those, 275,000 have come since the beginning of the Trump administration. What this indicates is not so much that AI has removed the need for human beings in the office, but that it has automated away the most inefficient paper-pushing jobs that only exist in taxpayer-funded entities that don’t have to worry about profitability.

This is bad news for government workers in the most bloated, useless positions but good news for taxpayers and legitimately overworked public employees. AI can make more objective hiring decisions, remove human biases from performance reviews, reduce workloads for public employees, and replace those who do little real work. All of these use cases provide better results at a lower cost to taxpayers.

We should all celebrate these innovations and welcome the possibilities. Instead, California union bosses are lobbying to keep government more bloated, less accountable, and more expensive for taxpayers.