# The State Must Never Own the Confessional

Companion Study Notes

## The Short Version

The relationship between artificial intelligence (AI) development, national security, and federal industrial policy is undergoing a significant transformation. By 2026, the U.S. government has moved from a traditional regulatory stance to a more interventionist approach, characterized by direct equity stakes in private companies. This shift is evident in the Pentagon's clash with Anthropic over ethical guidelines, the government's acquisition of nearly 10% of Intel, and OpenAI's proposal to offer a 5% stake to the administration to alleviate political pressures.

At the heart of this transformation lies a conflict between "constitutional AI," which emphasizes ethical constraints set by private companies, and the government's insistence on "any lawful use" of technology deemed critical for national defense. While supporters argue that these measures protect taxpayer interests and national security against adversaries like China, critics warn of potential risks, including the erosion of corporate fiduciary duties and the emergence of a "digital confessional" where sensitive citizen data could be accessed by the state.

## Why It Matters

This evolving landscape raises crucial questions about the balance of power between the government and private sector in the realm of AI. As the government takes on a more active role in equity ownership, concerns about privacy, corporate governance, and ethical standards intensify. The implications of this shift could redefine not only the relationship between citizens and the state but also the fundamental principles of corporate responsibility and individual privacy.

## Key Ideas

### 1. The Ethics-Sovereignty Conflict: Anthropic vs. The Pentagon
Anthropic has developed its "Claude" models with ethical safeguards that prevent misuse for mass surveillance or autonomous weaponry. This commitment to ethical AI has led to friction with the Department of Defense (DoD).

- **The "Supply Chain Risk" Designation:** After Anthropic declined to remove its ethical constraints for military use, Secretary of Defense Pete Hegseth labeled the company a "supply chain risk," potentially excluding it from defense contracts and discouraging partnerships.
- **The Pentagon’s Position:** Defense officials assert that AI funded by taxpayers must be available for "any lawful use," citing the need to compete with China's advancements in AI-driven military technologies.
- **The Constitutional Tension:** This situation raises a critical question: Can the government demand ethical responsibility from private innovators while exempting itself from those same standards?

### 2. The Expansion of Government Equity Stakes
The U.S. government has begun acquiring equity in numerous private companies in key sectors, moving beyond traditional grants and contracts.

| Company | Government Stake | Context/Value | Status |
| :--- | :--- | :--- | :--- |
| **Intel** | ~10% (9.9%) | Linked to $11.1B in CHIPS Act funding. | Finalized (subject to lawsuit). |
| **OpenAI** | 5% (Proposed) | Valued at ~$42.6B based on an $852B valuation. | Early-stage discussions. |
| **Chrysler** | 9.85% | Acquired during restructuring/TARP initiatives. | Historical/Active monitoring. |
| **General Motors** | 60.8% | Acquired during restructuring/TARP initiatives. | Historical/Active monitoring. |

- **The Intel Precedent:** Unlike the 2008 financial crisis, where government intervention was necessary to prevent insolvency, Intel's investment is framed as part of a broader industrial policy aimed at reinforcing domestic supply chains.
- **OpenAI's "Sovereign Wealth" Model:** CEO Sam Altman has proposed a model akin to the Alaska Permanent Fund, intending to share AI's financial benefits with the public while reducing regulatory burdens.

### 3. Regulatory Leverage and Allegations of Coercion
The line between regulation and ownership has become increasingly blurred, leading to legal challenges and claims of government overreach.

- **Extortion Allegations:** A March 2026 shareholder lawsuit (Paisner v. Tan et al.) alleges that the federal government used "extortionary threats" to pressure Intel into accepting its 10% stake, suggesting that the Department of Commerce hinted at "unpredictable consequences" for rejecting the deal.
- **Regulatory "Cudgels":** Critics argue that the administration employs heavy tariffs, export controls, and national security reviews as leverage. For example, Anthropic was compelled to restrict access to its latest models for non-U.S. citizens due to national security regulations, raising concerns about potential equity demands.

### 4. Privacy and the "Digital Confessional"
The government's ownership stakes in AI firms pose unprecedented privacy risks. AI interactions are increasingly seen as extensions of personal thoughts and emotions, akin to diaries or therapy sessions.

- **The HIPAA Gap:** Current HIPAA protections do not extend to general-purpose AI companies, leaving sensitive information about mental health, sexuality, and addiction vulnerable to exploitation.
- **The Threat of "Thought Policing":** If the government acts as both a shareholder and a regulator, it could gain access to private conversations, potentially enabling "thoughtcrime" policing through automated analysis of users' exploratory ideas.

## What To Listen For

As you engage with this topic, pay attention to the evolving dynamics between government intervention in the private sector and the implications for ethical AI development. Consider how the balance of power may shift and the potential consequences for privacy, corporate governance, and individual rights.

## Caveats / What Remains Uncertain

The long-term effects of government equity ownership in AI firms remain uncertain. Questions about the adequacy of existing privacy protections, the potential for regulatory overreach, and the implications for corporate fiduciary duties continue to be debated among experts and stakeholders. The evolving nature of this landscape necessitates ongoing scrutiny and dialogue to safeguard ethical standards and individual rights.