Tech 'Republicans' stage a messianic drama around AI

The author is the deputy editor of Content Division Three and the head of corporate research at the JoongAng Ilbo. In the early stages of the U.S.-Israeli war against Iran, Anthropic positioned itself against the White House and the U.S. Department of Defense, calling for limits on the weaponization of artificial intelligence. Recently, however, it has emerged as a central force in what could be called the "AI fear industry." After unveiling a new AI model in February that rattled software company stocks, the firm's founder Dario Amodei warned in a Fox News interview that within five years, half of all technical jobs could disappear, including entry-level roles for lawyers, consultants and finance professionals. Earlier this month, the company said its latest general-purpose AI, Mythos, was too dangerous for full public release, while providing preview access to about 40 organizations, including the U.S. government and major technology firms. According to foreign media reports, the U.S. National Security Agency is already using the Mythos preview. This follows an earlier designation of by the Pentagon of Anthropic as a supply chain risk entity, raising questions about whether the move was more symbolic than substantive. As OpenAI did after releasing ChatGPT, Anthropic appears to be using fear-driven rhetoric to establish a dominant position in the AI industry. The strategy is familiar: Warn that AI could rival nuclear weapons or destroy jobs, restrict access on the grounds of danger and then step forward as a rule-setter advocating tighter regulation. Even if parts of these claims hold merit, the broader pattern resembles a political act shaped by fear marketing. By emphasizing AI's risks, firms simultaneously present their own systems as essential safeguards. This elevates AI to a national security asset and positions its developers to influence state and even global decision-making. Those seeking a messianic role in the AI era raise a basic question: whose interests do they serve? The answer may be straightforward. Both companies are competing to go public this year, and few narratives attract investors more than war and job disruption. Korea, which has grown within a Western-led innovation ecosystem spanning semiconductors, the internet and AI, cannot ignore these developments. The challenge now extends beyond economics to politics and diplomacy, making the trajectory of such firms increasingly consequential. The case of Palantir Technologies, which supplied defense AI systems to U.S. forces during the war, offers a glimpse of future dilemmas. On Saturday, the company posted a 22-paragraph summary on X of its founder Alex Karp's book "The Technological Republic" (2025). The text reads as a declaration of technological patriotism, arguing that technology should serve to protect American power and values. It includes assertions such as the end of the nuclear age and the beginning of AI-based deterrence, calls to resist what it describes as hollow pluralism and suggestions that postwar military constraints on countries like Germany and Japan should be reconsidered. Palantir presents itself as an operating system for the U.S. government, but its ambitions extend across the broader Western alliance. As its role expands within a U.S. administration where governance itself is often questioned, allies such as Korea may increasingly face pressure to adopt these systems. Korea's push to develop an independent AI foundation model remains important for building its own ecosystem. Yet in practical politics, doubts persist about how much such efforts can achieve. The Korean government has already sought inclusion in the Mythos preview group, reflecting both dependence and urgency. As Silicon Valley's AI elite move onto the global political stage, Korea faces a difficult question: Can it maintain technological autonomy while upholding democratic values? The domestic political landscape offers little reassurance. The ruling party is ...