AIToday
Large Language ModelsAI Regulation & PolicyAI Safety & AlignmentFortune AIPublished: Jul 19, 2026, 19:00 JST2 min read

AI boom could entrench autocratic control, researcher warns

AI boom could entrench autocratic control, researcher warns

3 Key Points

  1. What happened

    A political scientist studying how autocracies respond to economic change argues that new AI sectors in countries like China are being structured to concentrate power around individual leaders rather than distribute it among independent elites.

  2. Why it matters

    The research suggests that when authoritarian regimes gain control over emerging industries—whether through tech commissions led by trusted allies or by disciplining independent entrepreneurs—they can create new power bases loyal directly to the leader, potentially weakening institutional checks and making repression more likely to protect those loyal networks.

  3. What to watch

    The author notes that China's Central Science and Technology Commission, led by one of Xi Jinping's closest allies, now oversees the AI sector, and the state is promoting AI firms aligned with Xi's priorities while disciplining more independent tech entrepreneurs—a pattern that may signal how autocrats will adapt to AI-driven economic shifts.

Ask the AI about this article →

Summaries like this, in your inbox every morning.

Context & Analysis

The article presents research on how authoritarian regimes adapt to structural economic change—a thesis built on a surprising historical pattern. The conventional wisdom, associated with Nobel laureates Milton Friedman and Friedrich Hayek, held that economic liberalization would weaken autocratic rule by making citizens less dependent on the state and therefore more capable of demanding political freedoms. Yet roughly half of all autocracies that opened their economies experienced increased human rights abuses instead. The researcher's answer centers on elite politics: when liberalization threatens the influence of regime insiders—party officials, military officers, and other power brokers—leaders must find ways to keep them satisfied. Repression serves this function by signaling commitment to protecting elite interests while allowing leaders to comply with international pressure for economic reforms. This creates a paradox where policies intended to democratize can actually entrench authoritarianism.

The AI sector now offers a new test case for this dynamic. In China, centralized control over AI—through a commission led by Xi Jinping's closest ally and through state backing of aligned firms—may be creating a new class of elites whose power depends on the leader's patronage rather than independent market success. If those new elites become an alternative base of support, they could reduce the leader's dependence on older institutional power centers, potentially creating space for greater centralization rather than diffusion of power. The researcher remains cautious about predicting the full trajectory, but the structural parallel is clear: emerging sectors of wealth and influence can be weaponized by autocrats to consolidate rather than distribute authority.

FAQ
How does the researcher connect AI to autocratic control?
The author argues that when AI sectors produce new elites whose success depends on their relationship with the leader—rather than independent business success—those elites become an alternative base of support for the leader, giving leaders more room to maneuver and potentially weakening institutional checks on power.
What specific steps has China taken to centralize AI control?
The Central Science and Technology Commission overseeing AI is led by one of Xi Jinping's closest allies, placing the sector closer to his inner circle. The state has been heavily involved in promoting AI firms aligned with Xi's priorities while disciplining more independent tech entrepreneurs.
What does the research predict about the outcome?
The author suspects the dynamics may fall in line with earlier patterns the book highlights, but notes: "I do not think we know yet how far that goes."

Get the latest Large Language Models news every morning

For example, today's edition would include:

  • Mastercard Launches Agent Pay for MachinesTop Companies AI · 23m ago
  • ServiceNow Becomes Core Partner in Multi-Agent AI PushTop Companies AI · 23m ago
  • LLM Security e-Learning Course Launches Oct 2026Top Companies AI · 23m ago

AI-summarized, only the topics you pick — one digest a day via Email, Slack, or Discord.

Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →

Ask AI

Ask AI anything about this article. Q&As are published on this page for other readers too.

Related Articles

Next articleAI Mania Distorting Corporate Decision-Making, Says Consultant