State Laws Fight for Classroom AI Safety as Federal Standards Lag

As states pass laws protecting students from classroom AI risks, a growing federal pushback threatens to override local student privacy and grading protections.

Wednesday, July 15, 2026

Key Takeaways

  • ## Illinois Restricts AI and Student Data Sales
  • Illinois bill SB3735 blocks educational software operators from selling student data. The legislation also restricts companies from training AI models on student records and allows parents to demand a human review of any work graded by an AI.
  • ## Federal Authorities Challenge State AI Laws
  • The federal government is actively challenging state-level AI regulations. To do this, the Department of Justice is using a litigation task force, while the Federal Trade Commission is issuing warning policies to state regulators.
  • ## Federal AI Safety Tests Ignore Classrooms
  • The federal Center for AI Standards and Innovation evaluates models from Google DeepMind, Microsoft, and xAI. However, these tests focus solely on national security. They do not look at classroom safety or child development.

State laws are stepping in to protect kids from artificial intelligence risks in classrooms, but federal opposition could soon wipe out these local rules. As school districts adopt more learning technology, parents and educators are caught in a battle over who controls student privacy and safety. While states push for strict local guards, federal agencies and tech companies are lobbying for a single national standard that would override local protections.

What Happened

State lawmakers are moving to regulate how technology companies handle student data and classroom tools. In Illinois, a legislative push under Illinois SB3735 targets educational software companies. If passed, the law would ban operators from selling or renting students' personal information or using that data to train AI models. The bill would also allow parents to request human review of any AI-generated grades and opt out of school-issued electronic devices.

Similarly, Virginia passed Virginia HB1186, which mandates that its state Department of Education establish guidelines for using AI in public schools.

At the same time, major AI developers like OpenAI are advocating for what they call "reverse federalism." In a recent policy post, OpenAI argued that states like Illinois, California, and New York are building a de facto national standard by passing similar risk-disclosure laws. However, OpenAI warns that a patchwork of state laws could slow down the rollout of safety tools, advocating instead for federal control.

The Bigger Picture

While states try to protect students, the federal government is working to curb state-level regulations. The Federal Trade Commission (FTC) recently proposed a policy warning that companies could face federal charges if they alter software output to comply with state laws, according to the Federal Register. The Trump administration has also launched a Department of Justice AI Litigation Task Force to challenge state-level regulations, as reported by Stateline.

In Congress, the bipartisan Great American Artificial Intelligence Act seeks to block states from regulating AI model development for three years, according to Global Policy Watch. While this bill would still allow states to regulate how AI is deployed in classrooms, critics fear it weakens the ability of local governments to protect citizens.

Meanwhile, the federal government's own AI safety system is not designed to protect children. The Center for AI Standards and Innovation (CAISI) recently signed agreements to test unreleased models from Google DeepMind, Microsoft, and xAI. According to Nextgov/FCW, these reviews are conducted in classified environments. These tests focus on national security threats, cyberattacks, and infrastructure risks. As a Forbes analysis points out, the federal testing framework ignores classroom concerns like developmental safety, age-appropriateness, and bias in automated grading.

What This Means for Families

This federal-state dispute leaves parents and educators in a difficult spot. If federal rules override state laws, local school districts could lose their power to regulate the tools used in their classrooms. The right to demand a human review of a child's automated grade or block biometric tracking could be stripped away by federal court challenges.

Because federal testing agencies like CAISI only evaluate models for national defense threats, schools cannot rely on federal approval to know if an AI tutoring program or grading tool is safe. Local administrators and parents must evaluate these systems themselves.

What You Can Do

  • Audit classroom tools: Ask your school district for a list of all AI-enabled software used in classrooms and check if they have a clear policy on human oversight for grading.
  • Review privacy agreements: Check if your child's school allows you to opt out of data collection, and verify that the district's technology providers do not use student work to train AI models.
  • Follow local legislation: Stay informed about state-level safety bills and contact state representatives regarding federal efforts to override local student privacy laws.
Share: