Explore how US government is championing digital freedom, fighting censorship, and shaping AI policy to preserve free speech while competing globally.
Digital Freedom, AI Regulation, and the Fight for the Western Internet
Key Takeaways
- Digital freedom is a national security imperative: As AI and foreign disinformation threats escalate, protecting internet freedom becomes critical to American competitive advantage and democratic values
- Government censorship apparatus is being dismantled: The State Department's new Digital Freedom Office reverses prior censorship efforts and prioritizes transparency, truth, and user empowerment over top-down content control
- Western AI must retain its soul: Preserving American legal frameworks like CDA 230 and fair use doctrine is essential to maintaining the US competitive edge in AI development
- European regulatory contagion threatens innovation: Vague, sweeping regulations about "risk assessments" and strict liability for AI outputs are spreading globally and could stifle creative development
- Rule of law must guide AI governance: Democratic deliberation through courts and legislative processes—not Silicon Valley executives—should determine controversial AI policies
- Viewpoint-neutral regulation encourages free speech: Government should create clear, consistent regulatory frameworks that don't weaponize rules against disfavored viewpoints
Understanding Digital Freedom as Public Diplomacy
Digital freedom has traditionally been viewed as a foreign policy tool to promote openness in authoritarian regimes. However, the conversation has fundamentally shifted. Under Secretary Sarah Rogers, who leads the State Department's public diplomacy efforts, has redefined digital freedom as both a national security issue and a core American value that requires active government support.
Public diplomacy extends far beyond traditional diplomacy between governments. It encompasses the relationship between the American government and foreign publics—including educational exchanges like the Fulbright program, media response capabilities, and critically, engagement with the information environment itself. This "operating system" of global communication determines how international conversations unfold and what narratives shape global perceptions of America.
The previous administration weaponized this machinery. The Global Engagement Center operated as a censorship apparatus, with State Department officials contacting Twitter and Meta to demand removal of content they deemed "disinformation"—including tweets from conservative commentators like Charlie Kirk. This apparatus became the subject of the landmark Murthy v. Biden Supreme Court case, which exposed government-social media collusion.
The Digital Freedom Office represents a complete reversal. Rather than suppressing speech deemed problematic by government officials, it now pursues transparency, truth, and reconciliation on prior censorship while making freedom of expression a primary pillar of American public diplomacy. This shift reflects a fundamental recognition: ** if America loses the information war, it loses its greatest soft power advantage**.
The Geopolitical Stakes: Why Internet Freedom Matters Now More Than Ever
The internet works in America. Citizens can text, post, share memes, and criticize government without fear of persecution. This freedom is not universal. In surprising countries across Europe, Asia, and the Middle East, internet access is increasingly controlled, monitored, and censored.
When the US claims to support internet freedom abroad, this commitment must be genuine and consistent. The problem is that American enthusiasm for internet freedom has historically been contingent—dependent on whether that freedom serves perceived American interests.
During the Arab Spring and Occupy Wall Street era, policymakers celebrated internet freedom because it empowered grassroots movements against "entrenched, closed, controlling institutions." The internet appeared to serve democratization. Then came 2016, Trump, and Brexit. Suddenly, the same policy establishment grew nervous about an "American Spring." Internet freedom—the very concept they had championed—began to look like a threat.
This anxiety triggered a massive pivot. Government internet freedom initiatives were transformed into disinformation curation programs designed to suppress what officials called "adverse narratives." The intention was prosocial: officials genuinely wanted citizens to have access to "true information" and protection from foreign adversary information operations, which are demonstrably real—especially in the age of AI.
But the remedy became worse than the disease. Bureaucrats overreached. They created opaque, upstream choke points far removed from user visibility. They funded NGOs to determine which arguments about sensitive topics (like pediatric transgender medicine) should be visible. They systematized censorship under the guise of fighting disinformation.
This represents a pattern that repeats throughout history whenever communication technology radically transforms. When the telegraph was invented, people feared it would destroy attention spans. When the printing press emerged, religious authorities panicked that heretics would print their own Bibles. When radio became ubiquitous, politicians worried about propaganda.
Today, we're living through a communications revolution exceeding all predecessors. The natural impulse is to "put the innovation back in the bottle"—to harness and control technology so legacy institutions can decide who benefits and what gets built. But this instinct is fundamentally inconsistent with American values and interests.
Dismantling the Censorship Apparatus: From Control to Empowerment
The State Department's reorganization marks a deliberate institutional pivot away from censorship and toward user empowerment. Under Secretary Rogers has acquired new authorities through the National Defense Authorization Act to promote internet freedom that her office previously lacked.
The Digital Freedom Office continues legitimate work that matters: combating malware, spyware, and foreign cyberattacks. But it has abandoned the problematic information integrity initiatives that mixed good intentions with authoritarian outcomes.
The new approach embraces transparency and user agency. Initiatives like ** content provenance tools** help people determine whether online content is authentic, AI-generated, or false—empowering users to make informed judgments. This is fundamentally different from the old model, where government-funded NGOs made determinations about acceptable arguments and invisible algorithmic filters suppressed disfavored viewpoints.
Similarly, the office now prioritizes censorship circumvention tools like VPNs, enabling people in repressive regimes to access information, and supports ** community-driven moderation systems like X's Community Notes**, where users transparently flag and provide context for potentially misleading content.
This represents the crucial distinction: tools and transparency that empower users versus ** opaque restrictions that concentrate power with institutions**. The former strengthens democracy and enables free speech. The latter, regardless of intention, recreates the censorious patterns America should oppose.
The AI Regulatory Crisis: Western Legal Frameworks Under Siege
AI represents the next frontier in this battle. The rules governing AI development are changing rapidly—and this is where American competitive advantage is most vulnerable.
American internet infrastructure rests on critical legal foundations often taken for granted because they're so foundational. CDA 230 is one such cornerstone. It allows platforms to publish third-party content without bearing the same legal liability as traditional newspapers. This seemingly technical provision is actually the structural bedrock that enabled the modern internet to exist. Remove it, and the entire ecosystem collapses.
Another pillar is the fair use doctrine, which permits transformative use of copyrighted material without permission. Courts have consistently ruled that training AI models on published works constitutes fair use—analogous to how a classroom of students learns from a library of books. This legal principle is essential to AI development.
These American legal innovations are now under assault—not by domestic critics, but by foreign regulators attempting to export their regulatory model globally.
The European Union, in particular, is advancing sweeping AI regulations that would devastate Western AI development. The problems are multifaceted:
Vague Liability Standards: EU regulations require companies to conduct "risk assessments" for hate speech, content that "adversely affects civil discourse," or content that might harm "well-being." These terms are undefined and indefinite. Does an AI system harm well-being if it's "too good" and people use it excessively? Does it violate regulations simply by being capable of generating certain content—even content that wouldn't be obscene under American First Amendment standards?
Strict Liability Regimes: Some EU draft legislation imposes criminal liability if an LLM is even capable of generating certain kinds of content. Anyone familiar with AI knows that these systems are probabilistic. Even with extensive safety architecture and careful training, LLMs emit unpredictable responses. Strict liability for capability creates an impossible standard.
IP and Transparency Overreach: While transparency is understandable, regulations requiring companies to disclose aspects of AI that reveal model weights enable foreign adversaries to reverse-engineer American systems, directly compromising Western competitive advantage.
Content Regulation Transplantation: The EU's content regulations for "very large online platforms" and "very large online search engines" could be transplanted onto AI systems. These regulations are notoriously vague, grant regulators enormous discretion, and create perverse incentives against innovation.
The danger is that global companies, facing potential fines up to 6% of global revenue, will comply with the most restrictive jurisdiction. This means European regulatory standards effectively become global standards, and American companies are forced to operate under rules designed by Brussels bureaucrats rather than American law.
This regulatory approach contradicts everything that made the US the engine of global innovation. It reflects the same impulse that created the censorship apparatus: a desire to control technology rather than liberate it.
Preserving AI's Western Soul: A National Security Imperative
Economist Tyler Cowen's concept of "AI with a Western soul" captures something essential. American AI systems are built on principles of individual liberty, democratic deliberation, and rule of law. These values are embedded in our legal frameworks and our approach to innovation.
There's a critical difference between American and authoritarian AI development. In China and Russia, AI is developed as a tool of state control—surveillance, propaganda, and suppression. In America, AI should serve individual freedom and democratic values. This isn't just ideological preference; it's strategic advantage.
When the US government contracts with AI companies through the Department of Defense or other agencies, the approach matters enormously. Critical decisions about AI capabilities—including controversial questions about autonomous weapons, data scope, and alignment—should be made through democratic deliberation, not Silicon Valley executive fiat.
This doesn't mean AI companies should lack autonomy. It means that rule of law should be the touchstone. Debates about what AI should do should occur in courts, legislatures, and public forums—the same crucibles of democratic deliberation that have served the Constitution for 250 years. These institutions are deliberately designed to be careful, principled, and consistent. They create precedent and transparency rather than arbitrary, changeable executive decisions.
The contrast is instructive. During the earlier censorship era, tech workers made decisions about acceptable speech with minimal democratic oversight. They determined, for instance, that it was impermissible on Twitter to call a convicted sex offender by their correct pronoun. These decisions reflect personal values—employees are entitled to their opinions—but they shouldn't replace constitutional governance.
This principle applies equally to AI. Tech workers and executives have a role, but ultimate authority over consequential questions should rest with democratic institutions and rule of law.
Building a Regulatory Framework That Encourages Free Speech
Elon Musk's acquisition of X demonstrated something important: a single visionary with resources can fundamentally reshape a platform's approach to free speech. By removing the "trust and safety" apparatus and replacing it with transparent, user-driven moderation (Community Notes), Musk showed that alternative governance models are possible.
But this is not a replicable playbook for most companies. Not every founder has Musk's resources or willingness to absorb the cost of upstream pressure from advertisers, governments, and activist groups.
So how can government encourage free speech without owning platforms? Several mechanisms exist:
First, abandon regulatory weaponization. Governments should refrain from creating regulatory cudgels that can be wielded arbitrarily against disfavored viewpoints. The EU's investigation into X's blue check verification system illustrates the problem: regulators threatened increased liability exposure if the platform allows President Trump to speak. This transforms ostensibly neutral regulations into viewpoint-based enforcement mechanisms. Similar patterns emerged in debanking cases, where financial regulations were enforced disproportionately against banks serving pro-gun groups.
This requires regulatory clarity and consistency. Laws should specify precisely what is required and what is permissible, leaving no room for arbitrary enforcement based on political preference.
Second, prioritize viewpoint neutrality in regulation. When government incentivizes content moderation, those incentives should encourage viewpoint neutrality. This doesn't mean platforms cannot moderate content at all. Spam, pornography, and other categories naturally receive different treatment because they generate negative user engagement. Providing tools to reduce spam, sexual content, or foreign disinformation from feeds is content-based moderation (appropriate), not viewpoint-based censorship (inappropriate).
The distinction matters. Viewpoint neutrality means regulators don't favor conservative speech over progressive speech, or vice versa. It doesn't mean platforms must host all content without distinction.
Third, provide legal certainty and protection. The US government should defend American companies when their interests and political freedoms are threatened by foreign regulators. If France threatened American platforms for hosting interviews with Emmanuel Macron, the US would rightly object to foreign interference in American speech. The same principle applies in reverse. When EU regulators attempt to control American platforms' speech policies, they're encroaching on American sovereignty and free speech values.
Fourth, shape international norms explicitly. As digital space becomes more important for commerce, international relations, and cultural exchange, the US must actively shape rules of the road. This requires both confrontational diplomacy (as Rogers has engaged in on X) and constructive bilateral/multilateral conversations. Secretary Rubio's recent speech emphasized that US-European alliance strengthens when allies are free, strong, and prosperous—which requires shared commitment to free expression.
The goal isn't to impose identical speech regimes globally. It's to prevent an "insidious, sweeping censorship contagion" where restrictions designed in Brussels spread globally and undermine Western values.
The Broader Principle: Freedom as American Strength
Throughout this conversation, a core principle emerges: freedom is not a liability to manage; it's America's greatest competitive and moral advantage.
During the Cold War, the US won ideologically because it offered freedom—economic, political, religious, expressive. Soviet systems offered control. The difference was stark and compelling.
Today's information competition mirrors this pattern. Authoritarian regimes are advancing AI, censorship, and surveillance systems designed to control populations. The US should compete by offering the opposite: AI systems built on liberty, platforms that enable expression, and legal frameworks that protect freedom.
This requires resisting the recurrent urge to control technology when it disrupts existing power structures. It requires trusting that freedom, transparency, and rule of law produce better outcomes than bureaucratic control.
Ironically, the evidence supports this trust. American internet companies—built on CDA 230 and fair use, powered by entrepreneurs with freedom to innovate—created technologies that served billions globally. The internet didn't emerge from careful government planning; it emerged from decentralized experimentation and legal freedom.
Similarly, American AI leadership exists precisely because founders can experiment, take risks, and build systems aligned with their values in a regulatory environment that permits innovation. If that environment changes—if regulations become vague, liability becomes strict, and enforcement becomes arbitrary—the incentive structure shifts. Companies become risk-averse. Innovation slows. Talent migrates to jurisdictions offering greater freedom.
This isn't theoretical. It's already happening. Companies are making decisions based on regulatory risk, and that risk is concentrated in the most restrictive jurisdictions.
Conclusion
Digital freedom, internet access, and AI governance are not peripheral policy issues. They are central to American competitiveness, national security, and the future of democratic societies globally.
The State Department's Digital Freedom Office represents a crucial institutional pivot—away from censorship and control, toward transparency and empowerment. Under Secretary Rogers' leadership, the US government is actively dismantling the apparatus that weaponized regulation against disfavored speech and building instead a framework that encourages free expression.
This work is essential because AI is becoming more important, not less. The rules governing AI development are being written now, in offices across Europe and America. If the US allows vague, sweeping European regulations to become global standards, Western AI loses its competitive edge and its soul. American legal frameworks—CDA 230, fair use, First Amendment principles—must remain foundational.
The path forward requires multiple actions: clear, viewpoint-neutral regulations; defense of American companies against foreign regulatory overreach; active diplomacy to shape international norms; and unwavering commitment to the principle that government should empower users and enable innovation rather than control speech and suppress dissent.
This is the fight for the Western internet. It's a fight for freedom—not as an abstract value, but as the concrete foundation that enables human flourishing, competitive advantage, and democratic governance. The stakes could not be higher.
Original source: Digital Freedom, AI Regulation, and the Fight for the Western Internet | The a16z Show
powered by osmu.app