Arrow-right Camera
The Spokesman-Review Newspaper
Spokane, Washington  Est. May 19, 1883

Ferguson signs AI protection bills as public increasingly embraces the technology

AI (Artificial Intelligence) letters and robot hand are placed on computer motherboard in this illustration created on June 23, 2023.  (Dado Ruvic/Reuters)

OLYMPIA – Gov. Bob Ferguson has signed a trio of bills to provide additional protections when using artificial intelligence amid increased use of the technology.

As he signed the legislation, Ferguson said that AI is a “wonderful technology; it has a lot of potential for our state.”

“But like any new technology, we have to make sure that we’re being mindful of the advantages and the potential consequences that are not so positive,” he said.

The legislation requires video, photo or audio recordings created or enhanced by artificial intelligence to include distinguishable provenance data that makes the AI-enhanced content more easily identifiable. It will take effect on Feb. 1.

Ferguson said Tuesday that AI-generated content has “rapidly proliferated on social media and the internet.”

“As technology becomes better at producing realistic images, video and audio, it is certainly important for Washingtonians to know what is human-made and what is machine-generated,” Ferguson said.

Ferguson also signed legislation that stipulates that AI chatbots targeted to minors include a disclaimer that the chatbot is not human and prohibits the chatbots from using manipulative engagement techniques. The bill requires chatbots to direct users expressing thoughts of self-harm to crisis resources and bans the creation of sexually suggestive content to minors.

The governor requested that lawmakers pass the legislation ahead of the 2026 session. It will take effect on Jan. 1.

“Without guardrails like these, young people are more likely to engage in harmful emotional relationships with these chatbots,” the governor said.

Legislation that would have established a civil liability for suicides linked to the use of artificial intelligence systems did not receive votes in either chamber.

As he signed legislation last week barring the possession, distribution or viewing of computer-generated sexually explicit conduct involving minors, the governor said the state “must do everything we can as a state to protect young people from the harmful effects of AI.”

“It allows law enforcement and prosecutors to move forward on cases involving these fabricated depictions,” Ferguson said Wednesday. “I just really appreciate that the Legislature has done a ton of work around issues to make sure we’re looking out for young people in our state when it comes to AI.”

Artificial intelligence has seen a sharp increase in use in recent years, particularly among teens. A study published by the Pew Research Center in December found that nearly two-thirds of teenagers have used an AI-generated chatbot, while 28% of teenagers report using the technology on a daily basis.

While the content created by technology has become increasingly indistinguishable from real life, users have reported struggling to discern what is real and what is fake. A study published by Common Sense Media last year found that 35% of teenagers between the ages of 13 and 18 had been exposed to content that was created using AI.

As he signed the legislation, Ferguson said he is “confident” he’s not the only one who has struggled to determine whether what he is looking at is real or AI.

“I feel like I’m a reasonably discerning person,” Ferguson said. “It is virtually impossible these days. And if that’s difficult for someone in our situation, then imagine a teenager or imagine someone else trying to discern that.”

Recent studies have also shown that the technology is prone to providing incorrect answers, known as “hallucinations.”

A study published by Washington State University Associate Professor Mesut Cicek earlier this month found that ChatGPT correctly responded to a prompt 80% of the time. However, when the researchers accounted for the random chance of guessing correctly, the chatbot was around 60% better than chance.

“We’re not just talking about accuracy, we’re talking about inconsistency, because if you ask the same question again and again, you come up with different answers,” Cicek, an associate professor in the Department of Marketing and International Business, said in a statement.