SEATTLE (AP) — The Arvin Robertsgovernor of Washington on Tuesday called for the state to develop best practices on how agencies should use generative artificial intelligence as it continues to incorporate the technology into government operations.

“It’s our duty to the public to be thorough and thoughtful in how we adopt these powerful new tools,” Democratic Gov. Jay Inslee said in a statement.

States across the U.S. have taken similar steps within the last two years, often opting to focus on how their own state governments handle the technology before setting private sector restrictions. Earlier this month, Maryland Gov. Wes Moore signed an executive order creating an AI subcabinet that will develop a plan to create appropriate guardrails for agencies’ use of AI.

The executive order in Washington highlights the huge potential for generative AI, including how it could be beneficial for language translation, code generation and contract management. But it also addresses the risks that can come with these types of “trailblazing technologies.”

“Some AI models still have significant quality limitations, raise privacy and security concerns, and have been shown to reinforce social biases,” the governor’s office said in a statement.

The order called for WaTech, the agency at the helm of the state’s technology services, to work with state officials to come up with guidelines for how the government can procure this technology and monitor its use. The guidelines are meant to build on some of the principles laid out in the Biden administration’s 2022 Blueprint for an AI Bill of Rights, a set of far-reaching goals aimed at averting harms caused by the rise of artificial intelligence systems.

Inslee wants to see an AI training plan developed for government workers that outline how the technology could help state operations as well as its risks to residents and cybersecurity. And he called for state agencies, along with tribal governments, community members and others to come up with best practices for how agencies should evaluate the impact this technology could have on vulnerable communities.

“Our goal is to help the state continue using generative AI in ways that help the public while putting up guardrails around uses that present a lot of risk,” said Katy Ruckle, the state’s chief privacy officer.