Oregon state agencies will soon have new safety standards under development for procuring and using advanced artificial intelligence systems. Governor Tina Kotek issued Executive Order 26-26 on Wednesday, directing the state's chief information officer to develop third-party safety review standards and assess whether so-called frontier AI models should be required to come equipped with a kill switch before Oregon government agencies can procure them.
Kotek Points Finger at Washington Gridlock
Kotek said state-level action was necessary because adequate federal safeguards are not in place, according to Oregon.gov. As KATU reports, Kotek said federal oversight and regulation would be strongest and most effective, but that states must act without adequate federal safeguards in place. She added that states must act urgently to manage how artificial intelligence is impacting their communities.
The order took effect immediately Wednesday. Per the same reporting, Kotek said Oregon should not rely solely on technology companies to establish their own safety standards, and that she remains committed to using her executive authority to put safeguards in place until the Oregon Legislature and Congress pass comprehensive standards for AI companies.
What the New Order Actually Requires
Executive Order 26-26, titled Establishing Responsible Artificial Intelligence Procurement Standards for State Government, directs Oregon to develop additional safety and oversight standards for procuring and using advanced AI models. The state's chief information officer must develop standards or criteria for third-party reviews of AI safety and must also assess the viability of requiring a kill switch mechanism for frontier AI models, the station's report notes.
That chief information officer, Terrence Woods, now has 90 days to develop and submit an implementation proposal back to the governor. Kotek has said she will reassess whether the order remains necessary every three months, and it will stay in effect until she terminates it.
Building on Years of Groundwork
Tuesday's order does not start from scratch. It builds on existing statewide requirements governing AI use across Oregon's executive branch, including the Oregon Responsible AI Usage Policy, which establishes governance for generative and agentic AI tools used by agencies, boards, and commissions, and already emphasizes risk management, human accountability, transparency, and oversight, per the same account.
That framework traces back to Executive Order 23-26, announced November 29, 2023, establishing an Oregon State Government Artificial Intelligence Advisory Council of no more than 15 members, according to Oregon.gov. Kotek announced the council's Final Recommended Action Plan on February 12, 2025, laying out five strategic focus areas covering governance frameworks, privacy, security, reference architecture, and workforce readiness. The recommendations drew on feedback from over nine committee meetings and 18 subcommittee meetings, alongside public input. Hoodline covered that milestone in our AI governance blueprint report last year.
Oregon's Broader AI Balancing Act
Oregon has offered AI training to state employees, according to OPB. In August 2025, the state named Nik Blosser, a former White House official and chief of staff to former Governor Kate Brown, as its inaugural Chief Privacy Officer and Artificial Intelligence Strategist, a dual role State CIO Terrence Woods created to lead enterprise data protection and workforce AI literacy, as detailed by StateScoop.
Oregon's order directs officials to develop third-party safety-review standards and assess whether frontier AI models should be required to have a kill switch.
