Governor Spanberger signed SB394/HB1186 into law on April 13, 2026, effective July 1, 2026, adding section 22.1-20.2:1 to the Code of Virginia. The law directs the Virginia Department of Education to publish guidance for the safe, ethical, and equitable use of artificial intelligence systems in instructional settings, and requires every school board to establish, implement, and enforce policies consistent with that guidance. The legislation also creates the AIS Innovation in Education Pilot Program to fund innovative AI applications in high-poverty, rural, and under-resourced divisions, with annual reports to education committee chairs and a sunset date of July 1, 2030. VDOE has already issued the 'Guidelines for AI Integration Throughout Education in the Commonwealth of Virginia,' covering guiding principles including do no harm, prioritize integrity, and augment rather than replace humans, as well as detailed roles and responsibilities for state agencies and educators [7]. In fall 2024, VDOE partnered with Germanna Community College and four additional community colleges to deliver statewide Generative AI workshops that trained 360 educators in AI-supported instructional strategies [8]. The VASS/VDOE Year of Learning initiative extended AI framework development to 75 school divisions [1].
Policy vacuum documented
Two research reviews published in the months before the law's effective date found that most Virginia school divisions have not yet adopted formal AI policies. The Metropolitan Educational Research Consortium at Virginia Commonwealth University examined governance documents from ten school divisions in June 2026 across six policy domains. Six of the ten had no AI-explicit language in their governance documents at all. Only four had some AI-explicit policy, with wide variation in scope. In the domains of staff use of AI and training and literacy, the infrastructure was weakest: two of ten divisions had explicit provisions for staff use, and three had them for training [1].
Separately, a George Mason University AI-LIFT review of 50 of Virginia's 131 school divisions, released in summer 2026, found that most divisions had not yet adopted a formal AI policy. Among the 23 rural divisions in the sample, only two had a formal AI policy. Of the 19 divisions with an adopted policy, fewer than half had also provided AI training to educators [4]. The report's URL path and internal date references indicate the data was current as of June 2026, though the precise public release date could not be independently confirmed.
National pattern of delayed adoption
The VCU/MERC report notes that the gap in Virginia mirrors a national trend. As of fall 2023, only 5 percent of U.S. school districts had adopted AI-specific student use policies, and approximately 90 percent of teachers reported that their district lacked policies addressing student AI use. The report also cites research showing that districts that invested in technology infrastructure and professional learning before the pandemic were better positioned for rapid adoption of AI tools, and that faster adoption in historically advantaged settings can widen existing disparities [1].
Data privacy and staff use gaps
The reviews point to specific vulnerabilities in data privacy. While all ten divisions in the VCU/MERC analysis had FERPA-aligned student records policies, nine of ten had no AI-explicit data protections. The report notes that FERPA was not designed for AI-specific risks such as model training on student data or agentic AI handling personally identifiable information. Only one of the ten divisions had addressed these questions explicitly in policy. The staff use gap is similarly wide: most divisions lack equivalent guidance governing what educators may do with AI tools. The VCU/MERC analysis found that only two of ten divisions had explicit staff AI use provisions, and that there is no technology policy equivalent to an acceptable use policy that reliably extends to staff AI use [1].
Implementation timeline ahead
The law sets no specific timeline for divisions to adopt policies, but creates a direct line of accountability from the state. The VCU/MERC report noted that the new law removes AI policy from the category of local discretion alone. VDOE has already begun the groundwork with training and framework development, but the research suggests that the majority of divisions are starting from a point of little existing governance [1].
