TAIPEI (Taiwan News) — Democracies can no longer rely on military strength alone to deter China, as Beijing is increasingly pursuing its strategic objectives through economic, technological, political, and information-based influence operations, a leading China expert said Tuesday.
Speaking at the China in the World Summit in Taipei, German Marshall Fund Indo-Pacific Managing Director Bonnie Glaser said China views competition with democracies as a comprehensive national strategy. She argued military strength alone is no longer a sufficient deterrent as China increasingly seeks to achieve its objectives through non-military means.
Citing research from Doublethink Lab’s China Index, which measures Beijing’s influence in more than 100 countries, Glaser said China’s influence is actually weakest in the military domain and stronger in foreign policy, technology, and the economy.
“China has made a strategic bet that it can achieve many of its objectives through non-military means,” she said. “Deterrence built only on military strength protects the domain where China is relatively less effective while leaving other critical domains exposed.”
Glaser said Chinese leader Xi Jinping’s concept of “comprehensive national security” has blurred the distinction between civilian and military sectors, domestic and international affairs, and even peace and conflict. “For Beijing, competition is no longer episodic. It is continuous, requiring persistent struggle across every domain of national power,” she said.
As a result, resilience should not be viewed simply as recovering from crises, Glaser said, but as the ability to deny adversaries the political, economic, and informational outcomes they seek. “When resilience changes the expected outcome of coercion, it becomes deterrence,” she said.
She added that national resilience depends on contributions from journalists, researchers, technology experts, universities, civil society organizations, and election officials as much as from the military.
“In the 21st century, deterrence is no longer built only at military bases,” she said. “It’s also built in newsrooms, research institutes, universities, election commissions, technology companies, and civil society organizations.”
German Marshall Fund Indo-Pacific Fund Senior Fellow Mareike Ohlberg said Beijing is incorporating AI into its domestic surveillance apparatus to process vast amounts of online data more efficiently and help authorities respond more quickly to perceived political threats.
She said Chinese companies working with government agencies are increasingly using AI for sentiment analysis, allowing authorities to determine whether online posts are positive, negative, or likely to spark public unrest. Ohlberg adds that AI enables China to make its censorship and information operations more scalable internationally.
As an example, China is already expanding its digital footprint overseas through the Digital Silk Road, a component of the Belt and Road Initiative. Chinese companies have signed agreements in countries including Kazakhstan, Uzbekistan, and Malaysia to build digital infrastructure and provide AI-related training.
Beyond surveillance, AI is also making Chinese information operations more effective by removing language barriers that previously limited overseas influence campaigns. “One of the bottlenecks China has faced is creating content in foreign languages,” Ohlberg said. “Artificial intelligence basically allows you to generate all that content automatically.”
She said AI enables influence campaigns to produce large volumes of tailored content in languages such as English, Spanish, Arabic, and French, making disinformation campaigns faster, cheaper, and more targeted. Ohlberg also warned that Chinese-developed large language models could export Beijing’s political narratives if they become widely adopted outside China.
She said free, open-source Chinese AI models may appeal to developers worldwide because they can be integrated into applications at little or no cost. However, models trained primarily on Chinese data are likely to produce responses consistent with Chinese government censorship.
“If more people internationally integrate those models into their apps, that is, of course, going to impact what comes out of that,” she said.
Rather than relying on a single solution, Ohlberg said democratic societies should strengthen media literacy, improve detection of coordinated bot networks, and expand “prebunking” efforts that warn the public about false narratives before they spread. She acknowledged that defending against AI-enabled disinformation remains a constant challenge akin to a cat-and-mouse game.
