Does a world run by AI scare you? WA lawmakers are trying to get on top of the tech explosion
When you hear the term “artificial intelligence,” what do you imagine?
Perhaps it makes you think of movies such as “The Terminator” featuring a murderous cyborg, or maybe a more friendly robot such as the title character in “Wall-E.”
Maybe it makes you think of the ways AI has been used in healthcare settings, such as transcribing medical records, diagnosing patients using large data sets, and even using AI to schedule doctor’s appointments.
Whatever version of AI comes to mind, there are a multitude more possibilities for its use. But while the number of valuable uses grows rapidly, people also are already using AI for nefarious purposes as well.
Just the other week, AI-generated depictions of pop star Taylor Swift performing sexual acts appeared on social media and spread like wildfire as thousands of accounts re-shared the photos. There is currently no federal law that regulates deep fake pornography, and only 10 other states have passed laws to ban it.
As AI becomes a bigger part of everyone’s day-to-day life, Washington lawmakers want to get in front of the advancements to put guardrails on the use of the technology.
Here are some of the policies that lawmakers are considering this session.
Fake intimate or sexually explicit images
HB 1999 is sponsored by Rep. Tina Orwall, D-Des Moines, and co-sponsored by 15 House Democrats and Republicans.
The bill would do several things.
First, it would create the offense of “Disclosing Fabricated Intimate Images,” committed when a person “discloses a fabricated intimate image of another person and knows or should know that the depicted person has not consented to disclosure, and that disclosure would cause harm to the depicted person.”
While an original version of the bill created new offenses in relation to crimes involving fabricated images, an amendment was added before passage out of committee that would expand existing criminal penalties under the current law on similar types of real images so that all prohibitions would be aligned.
The definition of “digitization” also was modified under the proposed substitute bill to include the “creation or alteration of visual or printed matter by using artificial intelligence.”
In addition, the bill also codifies a civil course of action for victims under the Uniform Civil Remedies for the Unauthorized Disclosure of Intimate Images Act.
Orwall told McClatchy in an interview that she and a team of other lawmakers began working on similar issues in 2015, including banning “revenge porn.” Last year, they proposed civil penalties for the release of intimate photos without consent.
Most recently, Orwall said she attended an event at the White House about intimate images where there was “a big discussion about AI and what that might do related to people being harmed.” Orwall’s bill was written based on those recommendations, she said.
Orwall said local incidents also inspired the bill, such as a high schooler in Issaquah who used photographs of female classmates to create nude AI-generated images of them. Because no laws exist to explicitly prohibit the practice, the student was unable to be charged with any crimes.
“I think, especially with this younger generation, they can be greatly harmed — they’re already at higher risk for mental health, for suicide — and it just really disturbs me to think that they also have to deal with such an invasive and harmful way to be attacked,” she said.
Because the images can be so realistic, Orwall said she believes they can be “just as damaging to a person.”
“Hopefully we can give a really clear message early on that there are ramifications of doing these kinds of harmful attacks,” Orwall said.
She said she believes there is still more work to be done, particularly in regards to AI-generated nude depictions of children.
The substitute bill passed out of the House Community Safety, Justice, and Reentry Committee unanimously on Monday, Jan. 29, and will head to the House floor for debate.
Sen. Mark Mullet, D-Issaquah, sponsored a similar bill this year in the Senate but that bill did not get a public hearing.
An artificial intelligence task force
HB 1934 would establish an AI task force to assess uses and trends, and to make recommendations to the Legislature regarding AI policy and regulation.
The bill is sponsored by Rep. Travis Couture, R-Allyn, and co-sponsored by House Republicans and Democrats at the request of the state Attorney General.
In an interview with McClatchy, Couture noted that AI is “powerful technology” and said that while Washington is “tech-centric” and wants to support new technologies, he thinks there should be societal guardrails.
Couture said that one of the biggest issues is that not many people truly understand what AI is, including some Washington lawmakers.
“You don’t want the legislature to have a knee-jerk reaction to a new technology, which could be used for good or evil, like any technology can,” he said.
The lawmaker said AI falls into two categories for him: risks and opportunities.
Risks, he said, include AI “revenge porn” or extreme bias in AI programs. Opportunities include agencies such as the Department of Natural Resources using AI to catch wildfires earlier.
As things stand, Couture said lawmakers are in “uncharted territory,” which is why the task force is so crucial.
“I think this step is important because it’s a task force filled with experts that will deliver non-binding solutions to the legislature about this,” Couture said. “And it could very well be that they deliver a lot of different things we need to dust off in the state law, or they can deliver very few things. That remains to be seen. But we should have that knowledge. And truly, this task force is late. If you really wanted to get ahead of AI, you needed this task force at least a year ago. But better late than never, in government terms.”
Couture’s bill passed out of executive session in the House Consumer Protection and Business Committee on Wednesday, Jan. 31, with three voting no.
A companion bill has been introduced in the Senate by Sen. Joe Nguyen. That bill was referred to Senate Ways and Means where it had a public hearing on Tuesday, Jan. 30.
Protecting employee rights
SB 6299 is sponsored by Sen. Derek Stanford, and co-sponsored by six other Senate Democrats.
The bill would prohibit employers from using a “replica of an employee’s voice or likeness without the employee’s explicit consent for a specific use.” The proposal would also prohibit employers from using AI to make employment decisions or evaluations without consent.
Stanford explained the bill further in a public hearing Jan. 29.
He said that although it is normal practice now for employees to consent to be in pictures taken at the workplace, “it’s very different in a world where you can take pictures of someone and then use their likeness in an AI-generated model.” That needs to be a different discussion and made clear to the employee, he said.
Additionally, Stanford said, the bill would affect “disclosure around the use of AI processes in evaluation of employees, and potential employees.”
“I don’t think it’s asking too much to have a little bit of disclosure so people know what’s going on,” Stanford added.
The bill passed out of the Senate Labor and Commerce Committee on Jan. 30 with only Democratic support.
State guidance on AI usage
State House and Senate lawmakers aren’t the only ones trying to find ways to navigate the AI wild west.
On Jan. 30, an executive order was signed by Gov. Jay Inslee to direct state agencies to adopt policies surrounding the ethical use of AI.
“This executive order lays out a year-long process for agencies working together to assess the feasibility, benefits and challenges of integrating this technology into agency operations and services,” Inslee said in a news release Jan. 30. “It’s our duty to the public to be thorough and thoughtful in how we adopt these powerful new tools.”
WaTech is the agency in Washington that operates state technology and information security services. Under the order, they are tasked with working with other executive branch agencies to identify “potential generative AI initiatives that could be implemented in state operations.”
The agency is then responsible for submitting a plan to “develop initial guidelines for how the government may procure, use and monitor the use of generative AI.”
Inslee is the 10th governor to issue an executive order regarding AI, and Washington’s order is similar to the order issued in California by Gov. Gavin Newsom.
This story was originally published February 1, 2024 at 5:00 AM.
