Minister of Digital Affairs Lin Yi-jing (林宜敬) recently visited Washington and met with US officials, think tanks and major technology companies to discuss cooperation on artificial intelligence (AI) and cybersecurity, his ministry said yesterday.
During the trip, Lin and his delegation held talks with US government officials, who the Ministry of Digital Affairs did not identify, and several Washington-based think tanks, including the Hudson Institute and the Center for Strategic and International Studies.
One of the main issues discussed was “how AI models understand Taiwan,” as the government seeks to ensure the country’s language, history and culture are accurately represented by AI systems.
Photo courtesy of the Ministry of Digital Affairs via CNA
The issue is no longer merely a technical one, because growing numbers of people are turning to AI for information and to make sense of the world, Lin said.
Taiwan is developing what it calls “sovereign AI” capabilities using traditional Chinese and local data, the ministry said.
Sovereign AI broadly refers to efforts by countries to gain greater control over the infrastructure, data and expertise underpinning their AI systems.
The government is also evaluating how international AI models handle Taiwan-related issues and sharing the results with developers to help improve their models, the ministry said.
“Some authoritarian countries” are attempting to use AI to reinterpret history and influence how societies develop, Lin said.
That makes it important for Taiwan to preserve its own data, language and perspectives, and ensure the perspectives of democratic countries such as Taiwan and the US are “properly understood in the AI era,” he said.
Cybersecurity also featured prominently in the discussions.
Taiwan faces about 2.6 million cyberattacks per day and had recently observed hackers using AI to make attacks more efficient, the ministry said.
The Taiwanese delegation discussed cyberdefense, threat intelligence sharing, scenario exercises, post-quantum cryptography and frontier AI with its US counterparts.
They also shared Taiwan’s experience in responding to cyberthreats.
Taiwan hopes to deepen exchanges with the US on advanced AI models and technologies, and explore their use in reinforcing cybersecurity at government agencies and critical infrastructure, the ministry said.
During talks with OpenAI, the company behind ChatGPT, the delegation discussed using advanced AI for threat analysis, vulnerability detection and incident response.
Discussions with Meta, which owns Facebook and Instagram, covered scam prevention, AI-generated content and platform governance, including ways to improve age and identity verification while protecting user privacy.
The delegation also held talks with Amazon on satellite communications and the development of related services in Taiwan. Amazon operates Amazon Leo, a low-Earth orbit satellite network.
