Hawaiʻi’s 2026 Legislature faced more decisions about artificial intelligence than any session before it, touching government services, health care and children’s lives. Some passed. Many didn’t. The gap between the two says a lot about our approach.

Artificial intelligence was among the most contested issues in state capitols this year. According to the National Conference of State Legislatures, 1,263 measures to address it were introduced across 48 states in 2026. President Trump tried to preempt these with an executive order and funding threats, but lawmakers from both parties pushed back.

Child safety was the exception and the area with the most agreement. Hawaiʻi’s new law requiring AI companion apps to disclose they are not human, maintain protocols to address self-harm, and protect minor users was one of at least six similar laws enacted across the country in 2025 and 2026. It is the clearest example of what this session did well. The rest is more complicated.

A Special Commentary Project

Civil Beat is focusing on transparency, accountability and ethics in government and other institutions. Help us by sending ideas and anecdotes to sunshine@civilbeat.org.

Read More

Beyond child safety, lawmakers targeted direct harms by prohibiting AI-generated deepfakes and creating civil remedies for unauthorized disclosure of intimate images. The Legislature also addressed digital deception, banning government-impersonation scams, requiring social media platforms to allow permanent account deletion, and mandating that hotels and ticket sellers display total prices up front.

What the session did not produce was any structural framework for AI governance. No protections against algorithmic discrimination. No requirement that developers disclose how their systems work. No guaranteed right to human review when AI makes major decisions. While states like Colorado, Utah, California, New York and Texas have moved toward broader governance frameworks, Hawaiʻi’s bills were all deferred in committee.

Without a clear set of rights or even guidance for specific industries, artificial intelligence systems will have too much power in our day-to-day lives. For example, an insurer in Hawaiʻi can use an algorithm to deny your claim without human review or disclosure. More than a dozen other states have passed laws requiring oversight, but similar bills in Hawaiʻi failed to advance.

Surveillance pricing, the practice of using customers’ personal data to charge different shoppers different prices for the same item, is on the legislative agenda in at least 11 states this year. Maryland became the first to ban it for food retailers this spring. Hawaiʻi also considered a bill this year but it was deferred in its first committee. Lawmakers should return it to the agenda next session.

Children’s access to social media is another issue that isn’t going away. Bills that would have required age verification and parental consent before minors could download social media apps never moved out of committee, even as at least 19 states enacted restrictions and Australia banned children under 16 from platforms entirely. The hesitation has a logic. In the purpose clause of the companion app law, legislators explicitly warned against requiring “identity documentation for age verification purposes” as a data collection risk.

Phone bans in schools present no such trade-off. Multiple bills that would have required the Department of Education to prohibit phones during the school day never moved out of committee, even as states and countries around the world have enacted similar restrictions.

Another of the most hotly contested bans nationally didn’t come up here at all. Lawmakers in at least 11 states introduced bills to temporarily halt data center development, driven by concerns about rising electricity bills and environmental impacts. Maine came closest, passing a moratorium that the governor vetoed. Hawaiʻi hasn’t reached the same level of demand as other states, but it’s growing. Maybe this session wasn’t the moment, but the next one could be.

That uncertainty is not unique to Hawaiʻi. The technology is moving faster than any government, federal or state, has figured out how to manage. Every legislature grappling with these issues has made similar calls: which harms are concrete enough to act on, which trade-offs are worth making, which fights to defer. States are testing approaches, watching what other states do, and revising. Hawaiʻi is part of that process.

But lawmakers can’t only focus on regulating and restricting technologies. This session, legislators also moved to authorize mobile driver’s licenses, expanded automated speed cameras to high-risk locations, allowed courts to impose intelligent speed assistance systems on repeat offenders, and modernized emergency services to include telehealth and mobile health care. New technologies create opportunities to improve government services.

That said, we have to implement them responsibly. Some states that have deployed AI to flag benefits fraud have gotten it badly wrong. In Michigan, an automated system wrongly accused 48,000 people, and a pandemic-era audit found 93% of fraud flags were false. Yet, the federal government is still pressing states to automate income and work verification for programs like SNAP and Medicaid.

Hawaiʻi will face that pressure too. What this session showed is that the Legislature can act on both fronts: restricting technologies that cause specific harm and adopting technologies that improve government services. Both are necessary. Neither is sufficient on its own.

The gap between what passed and what didn’t says something about our approach. We can react to problems in front of us, but the test for next session is whether Hawaiʻi can get ahead of what’s coming, not just respond to what’s already here.