Washington state Sen. Lisa Wellman is a self-described science fiction fanatic.
Wellman regularly gives teenagers Isaac Asimov’s 1950 short story collection I, Robot. She recalls the book’s famous “Three Laws of Robotics,” which include the provision that “a robot may not injure a human being or, through inaction, allow a human being to come to harm.”
It’s a good lesson, Wellman said, one that some in the tech industry would do well to remember.
Wellman, a Democrat from Bellevue, is the sponsor of Senate Bill 5984, which aims to protect young people from the negative mental health effects of artificial intelligence chatbots. It’s one in a slew of AI-related bills introduced by state lawmakers this year.
“I have not seen what I would call responsible oversight in products that are being put out on the market,” said Wellman, who spent decades in the tech industry working for Apple before becoming a lawmaker. “I’m not even sure that the developers that put the products out in the market have a clear understanding of what they’re capable of.”
Washington state doesn’t have any major AI regulation on the books. But lawmakers are increasingly concerned about the technology’s potential harms, and are looking to catch up.
“I think we’re all very conscious of the issue,” Wellman said.
While many of this year’s proposed AI regulations appear dead for the session, several major bills are advancing, and lawmakers say they feel a sense of urgency. In addition to Wellman’s chatbot bill, lawmakers are looking at legislation related to data centers, AI surveillance in schools, AI-generated content detection and laws to give people the right to their own AI-generated likeness.
“This year, in my mind, is kind of the beginning of trying to address this great leap of artificial intelligence,” said Rep. Clyde Shavers, a Democrat from Oak Harbor who sponsored the bill about AI content detection.
The tech industry has lobbied against many of these proposed AI regulations, arguing that they are too broad and could stifle progress by opening companies up to excessive liability.
“There are concerns about, if you do overregulate, losing access to certain tools, or losing the full capability of AI tools,” said Aodhan Downey, the western state policy manager for the Computer and Communications Industry Association, a trade association that includes Google, Meta, Amazon, Waymo and other large tech companies.
Lawmakers such as Wellman disagree, saying that the industry has been allowed to run unchecked for too long.
“We’re looking at things that have the power to do real, incredible damage,” Wellman said.
Chatbots
Over the past few years, a growing list of lawsuits and news reports have linked instances where young people died by suicide to extended interactions with chatbots. The topic “Deaths linked to chatbots” now has its own Wikipedia page.
The trend has alarmed lawmakers in Washington — and across the country. Lawmakers in at least 27 states introduced legislation to regulate chatbots this year.
The issue is also a major concern for Gov. Bob Ferguson, a parent himself. He requested the chatbot bill and listed it as one of his top priorities this session.
House Bill 2225 and its companion, Senate Bill 5984, would require that chatbots give users hourly reminders that they are talking with an AI, not a real human. The legislation would also require developers to implement suicidal ideation detection and prevention protocol; regularly report data; and take “reasonable measures” to prevent the chatbot from showing explicit content to minors or mimicking a romantic relationship with them.
The legislation relies on private right of action tied to Washington’s Consumer Protection Act, meaning anyone can sue if they feel a company isn’t compliant. That aspect has been met with pushback from the tech industry. Downey, with the technology trade association, worries it would open up developers to “additional liability that they probably aren’t comfortable with.”
AI detection
State lawmakers are also worried about AI-generated content making it difficult for people to tell what’s real.
House Bill 1170 would require AI developers to include a disclosure — such as an embedded watermark — in AI-generated images, audio and video to allow users to detect if something was generated or altered by AI.
“It tells AI systems: ‘If you generated it, or if you altered it, the public deserves clear notice,’” said Rep. Shavers, who sponsored the bill.
Shavers stressed that his goal is to improve consumer protection while ensuring that innovation isn’t hindered.
But some in the tech industry have pushed back, saying that the bill contains overly broad definitions and that implementing it isn’t feasible. During a recent Senate hearing, Amy Harris, director of government affairs for the Washington Technology Industry Association, argued that the bill is “unworkable” and contains definitions that conflict with other states’ laws.
Both the chatbot bill and the AI detection bill are modeled after legislation that passed in California. As Washington and other states consider similar AI regulations, some developers are concerned about navigating a maze of overlapping but slightly different laws.
“Being able to come into compliance with a bill in one state that puts you out of compliance with a bill in another is just an impossible place to be,” Downey said.
Last year, President Donald Trump pushed for a 10-year moratorium banning states from adopting their own AI regulations. The proposal didn’t move forward, but lawmakers say the lack of federal action has added urgency to their efforts to regulate AI at the state level.
Data centers, surveillance, and digital likeness
Several other AI-related bills are moving forward this session.
Senate Bill 5956 aims to protect students from AI-driven discipline and surveillance in schools. It would prohibit school districts from using automated decision systems, facial recognition or biometric data as the sole basis for disciplinary decisions. It also prohibits things like predictive risk scoring and biometric surveillance.
Senate Bill 5886 would expand Washington’s right-of-publicity law to cover an individual's AI-generated digital likeness. It would prevent an AI depiction of someone's voice or image from being used for commercial purposes without their permission.
Lawmakers are also looking to regulate data centers. The rise of AI has led to a boom in power-hungry data centers. Data centers are expected to be the biggest drivers of electricity demand in the Pacific Northwest in coming years, and have prompted concern about environmental impact and higher electric bills for consumers.
House Bill 2515 would require utilities to set a tariff or policy to prevent financial risk to ratepayers; submit annual sustainability reports; and restrict data centers from receiving no-cost allowances under Washington’s Climate Commitment Act.
Some of this year’s AI proposals failed to advance, including a proposal to regulate “high risk” AI systems and protect people from algorithmic discrimination; restrictions of algorithmic price setting; new requirements around AI training data transparency and additional consumer protections for AI systems.
If you or someone you know is contemplating suicide, call for help now. The National Suicide Prevention Lifeline is a free service answered by trained staff. The number is: 1-800-273-8255.
All stories produced by Murrow Local News fellows can be republished by other organizations for free under a Creative Commons license. Image rights may vary. Contact editor@knkx.org for image use requests.
