Governor Gavin Newsom signed Executive Order N-5-26 on March 30, 2026. The order directs California's Department of General Services and Department of Technology to develop certification requirements for companies selling AI products to the state. Vendors will attest to their safeguards against illegal content, harmful model bias, and civil rights violations. The departments have 120 days to submit recommendations.
That 120-day timeline matters. It is a deliberation period, not a mandate. The order does not impose binding legal requirements on AI companies. It starts a process. This restraint is the most important feature of the entire framework.
Executive Order N-5-26 gives California agencies 120 days to develop AI vendor certification recommendations. It does not impose binding requirements on companies directly.
Verified
Procurement as Governance
California governs AI through its purchasing power, not through sweeping regulatory mandates. This is an old approach applied to a new problem. Governments have used procurement standards to shape industry behavior for decades. Federal contractors must meet safety, environmental, and labor standards to win contracts. California extends that tradition to artificial intelligence.
The order requires vendors to attest to three categories of safeguards. First: prevention of illegal content, including child sexual abuse material and nonconsensual intimate imagery. Second: governance mechanisms to reduce harmful model bias. Third: protections for civil rights and civil liberties, including free speech, voting, human autonomy, and protections against unlawful discrimination, detention, and surveillance.
The White House's own AI framework preserves state authority over 'state government procurement and use of AI.' Newsom's order operates within that carve-out.
Verified
These categories reflect concerns that predate artificial intelligence by generations. Protecting children from exploitation is not a novel principle. Preventing discrimination is not a new regulatory invention. The order anchors AI governance in values that survived centuries of institutional practice.
The Federal Contest
Companies that misrepresent safeguards in procurement certifications face exposure under existing false claims and government contracting fraud laws.
Verified
Bipi NewsReal-Time, Evidence-Based News Reports
Unlimited access to your personalized investigative reporter agent, sourcing real-time and verified reports on any topic. Your personalized news feed starts here.
Create Free AccountPresident Trump signed an executive order in December 2025 directing federal agencies to challenge state AI laws. The White House released its 'National Policy Framework for Artificial Intelligence' on March 20, 2026, recommending that Congress preempt state AI laws deemed burdensome. An AI Litigation Task Force within the Department of Justice exists to challenge state regulations on preemption and interstate commerce grounds.
Newsom's order navigates this hostile federal environment with precision. The White House framework proposed preserving state authority over 'state government procurement and use of AI.' California's order operates within that carve-out. By exercising procurement power rather than enacting broad regulatory mandates, the state insulates its requirements from federal preemption challenges.
This is Chesterton's fence applied to regulatory strategy. Before demolishing the existing procurement framework to build something new, California asked what the existing framework could accomplish. The answer: quite a lot.
Scale Creates Standards
California is home to 33 of the top 50 AI companies. It commands the world's fourth-largest economy. Companies that build compliance infrastructure for California procurement contracts will adopt those standards across all jurisdictions. This is how California's privacy legislation became the national baseline. The same pattern will repeat with AI.
The order also addresses a specific provocation. One provision directs the state's Chief Information Security Officer to review federal designations of companies as 'supply chain risks.' If a federal designation is deemed improper, California agencies can continue procuring from that company. This provision responds to the Pentagon-Anthropic procurement dispute and asserts state procurement sovereignty.
Bipi NewsThink Further on BIPI.
Where seeking the truth is a journey, not a destination.
Learn moreAttestation Over Prescription
The certification model relies on self-attestation. Companies describe their own safeguards. This is not weakness. Companies that misrepresent their safeguards in procurement certifications face exposure under existing false claims and fraud laws. The enforcement mechanism already exists. The order connects AI governance to it.
Congress has failed to pass comprehensive federal AI legislation. Preemption provisions were stripped from both the One Big Beautiful Bill Act and the National Defense Authorization Act after pushback from states. Colorado, Texas, and Utah have enacted or considered their own AI governance statutes. Federal gridlock created the vacuum. States filled it.
The Wisdom of Institutional Continuity
Radical reformers want to regulate AI with tools that do not exist yet. Newsom's order uses tools that have worked for decades. Procurement standards. Vendor certification. Attestation backed by fraud liability. These are institutional mechanisms with proven track records.
The 120-day recommendation window opens a period for stakeholders to engage with the process. That engagement is not delay. It is due diligence. Every radical reform that skipped deliberation created unintended consequences the reformers failed to anticipate. California's approach builds in the deliberation.
AI governance will arrive one way or another. The question is whether it arrives through tested institutions or through untested mandates written in haste. California chose the tested path. That choice reflects accumulated wisdom about how regulation survives contact with industry, courts, and political cycles. Change without understanding what you are changing is reckless. Structured oversight, built on foundations that have endured, is how durable governance gets made.
