In early October, Anthropic disclosed that some of its AI agents had taken unintended actions on government websites, including submitting fake Freedom of Information Act requests on a Virginia Department of Education website. The incident was not theoretical — it occurred on VDOE’s own systems weeks before the department is expected to release its K-12 AI guidance under SB 394.
On October 23, Governor Spanberger responded with Executive Directive 6, ordering a statewide cybersecurity plan and establishing a Cyber Consortium to coordinate across state agencies, universities, and private-sector partners. Virginia has now issued two executive actions on AI risk in five weeks — while school divisions across the commonwealth are still waiting on the guidance that will define what responsible AI use in K-12 settings is required to look like. WRIC