Governments Should Regulate AI Misuse and Harm, Not the Underlying Technology – Capitol Weekly | Capitol Weekly

OPINION – As a former corporate employee and current minority business owner, I have worked tirelessly to reduce bias and discrimination in the workplace. At my new company, we are successfully leveraging artificial intelligence to solve these thorny problems. So I was surprised and disappointed when I recently saw a well-intentioned, but terribly misguided, AI bias bill being considered by the California Legislature.

In 2021, I started building an AI-powered coaching platform that helps strengthen emotional intelligence and soft skills in the workplace. This helps companies and individual employees in two important ways: First, it helps people recognize if bias is lurking in their thinking and decision-making. Second, it helps people with bias-related concerns or frustrations to articulate them in effective communication that is more likely to generate constructive responses rather than defensive ones.

Helping employees create positive change is empowering for them and good for companies. As a Latinx manager early in my career, I wish I had had similar mentorship, and I’m proud that our young company has innovatively used AI to help so many people and employers achieve this. I hope California lawmakers stop rushing to enact AI into law and appreciate its positive power.

Most of the discussion on AI in Sacramento seems to be centered on Senator Wiener’s bill, which focuses on large-scale AI models and the potential risks. But Bill 2930, authored by Representative Rebecca Bauer Kahan, frustrates me even more because, while I fervently support its anti-discrimination goals, I know it will undermine our ability to reduce bias in the workplace.

Anti-discrimination laws have historically operated retroactively because whether a decision was discriminatory or what factors produced a discriminatory outcome can only be determined after the decision was made, or after many decisions have been made and patterns become apparent. AB 2930 is a noble attempt to prevent discrimination prospectively by regulating technology before it is used. While the proposed regulations would not mandate how AI models or AI-influenced services should be built or operated, they would make pre-release risk assessments, predictions, evaluations, and mandatory governance programs costly and difficult for all AI startups like mine.

I know many founders who have built AI-leading companies that are actively tackling the problems this bill seeks to prevent — using AI to help people get better education, personalized healthcare, and lower-cost mortgages. But if AI startups are burdened with costly governance obligations, burdensome risk assessment and reporting requirements, and constant evaluation mandates, the funding will dry up and the mission will be aborted.

In contrast, the government does not impose these burdens on companies that make similarly critical decisions using only their brains and computers, or that receive decision support from consultants. Why would the government bury small businesses in paperwork that use technology responsibly to make better decisions? If AB 2930 becomes law, startups like mine will be at a significant disadvantage and large AI developers will have an advantage. The ability of small businesses to build specialized AI solutions that solve addressable problems and contribute significantly to California’s world-leading innovation economy will be severely hampered.

California has a long history of strong anti-discrimination laws that I support, and I’ve dedicated my career to helping people feel safe and supported in the workplace. Without the magic of AI, this would be impossible to achieve at scale, and it certainly will not be if AB 2930 passes.

I’m a proud AI optimist and won’t shy away from the risks, but I urge lawmakers to reject legislation driven by fear and embrace AI’s opportunities. We need balanced legislation that promotes the benefits of AI and addresses the risks. Lawmakers must do more.

Remy Meraz is the co-founder and CEO of Zella Life, an AI-powered coaching platform that helps strengthen soft skills and emotional intelligence in the workplace.