Understanding Expressive Governance: The Intersection of AI, National Security, and Free Speech
The realm of artificial intelligence (AI) is rapidly evolving, raising complex questions about ethics, safety, and governance. Recently, a prominent AI company faced severe repercussions when it refused to allow its technology to be utilized for autonomous lethal weapons or mass surveillance of citizens. Rather than seeking an alternative vendor, the U.S. government branded this company as a “supply-chain risk to national security,” a designation typically reserved for foreign adversaries. This situation prompted a broad administrative response, including an order from the President for all federal agencies to cease using its technology.
What Is Expressive Governance?
In a forthcoming book and an amicus brief filed in the case of Anthropic v. U.S. Department of War, the concept of expressive governance is explored. This term describes the executive branch’s strategic use of standard administrative tools—such as funding, contracting, and regulatory designations—to penalize dissenting viewpoints while maintaining an appearance of legality. This “punishment for protected speech” raises critical questions about the limits of administrative discretion and its implications for free speech.
The Dangers of Expressive Governance
The primary concern with expressive governance lies in its façade of legality. Courts often defer to the executive’s discretion regarding contracts, yet this deference is intended for genuine management decisions, not retaliatory actions. The implications are alarming; when the government activates these low-visibility levers to suppress dissent, the constitutional violation can remain obscured within the bureaucratic process.
This form of governance extends beyond this single case. It affects law firms stripped of access for representing unpopular clients, universities threatened with funding loss over faculty speech, news organizations facing regulatory reprisals, and civil servants dismissed for dissenting opinions. Such actions have already drawn judicial condemnation as unconstitutional, exposing the tactic’s common foundation of neutral-seeming authority inflicted on those unwilling to comply.
A Chilling Effect on Free Speech
As Luke Barnes pointed out, the repercussions of this governance method send a clear but unsettling message: industries reliant on federal contracts should suppress any principled stands against the current administration, lest they become targets themselves. This creates a chilling effect—where the fear of reprisals discourages speech and positions that the First Amendment aims to protect.
More Read
![Exploring Unilateral Revision Power in Human-AI Companion Interactions: Insights from Research [2603.23315]](https://aimodelkit.com/wp-content/uploads/2026/04/1776481243_arxiv-logo-fb-150x150.png "Exploring Unilateral Revision Power In Human-Ai Companion Interactions: Insights From Research [2603.23315]"){kind=logo}
Legal Recourse and Judicial Intervention
The challenging question remains: how can courts effectively address expressive governance? Striking down specific orders at a time may yield minimal results when such tactics can quickly resurface in new forms. To combat this, a structured framework could establish a repeatable method for judges, as proposed in the Anthropic case brief.
The Proposed Framework
-
Clear-Statement Rule: Before the executive can use discretionary powers that impact speech, there should be a requirement for a clear law that authorizes such actions. Stretching designations meant for foreign threats to punish domestic companies requires explicit Congressional guidance, which is absent in cases like this.
-
Burden of Explanation on the Government: Once a plaintiff demonstrates a pattern linking action to viewpoint—highlighted by timing, targeting, or public statements—the government must prove that its decisions would have been made for legitimate reasons. This approach prevents the concealment of retaliatory motivations behind bureaucratic pretexts.
-
Presumption on Expressive Intermediaries: Courts should apply the strongest presumption when the government targets what are termed “expressive intermediaries.” These entities significantly shape public discourse, and retaliating against them affects communication on a broader scale. In cases involving AI providers, for instance, their articulated safety principles serve as their speech, influencing how numerous individuals and organizations process information.
Preserving the Integrity of Free Speech
This proposed framework does not create a new tier of scrutiny or restrict executive powers. The government retains the right to disagree with a company, reject propositions, and seek alternative vendors. However, weaponizing national security to undermine a business for its viewpoints crosses ethical boundaries that courts should be equipped to identify.
The stakes extend far beyond individual companies or administrations. If expressive governance prevails, the cost of voicing one’s beliefs becomes ruinous, steadily constricting the domain of authentic public discourse. Courts, positioned uniquely in this landscape, must actively intervene to uphold the tenets of free speech and democracy.
Inspired by: Source
