Taiwan rolls out draft artificial intelligence law

Taipei, July 16 (CNA) The Cabinet-level National Science and Technology Council (NSTC) unveiled on Monday a draft proposal for an artificial intelligence (AI) law that covers the use and reliability of AI and mitigation of the risks associated with it.

The draft law drawn up by the NSTC would require the government to establish a framework delineating responsibility for AI applications consisting of labeling, disclosure, and accountability mechanisms to strengthen the trustworthiness of such applications.

According to the proposal, however, norms on responsibilities for AI applications would not apply to any activities in the AI technology development and research stages prior to the release of an application to support technological innovation.

The law would also stipulate that additional regulations governing the use of AI applications, AI accountability, or how AI can be used in the insurance industry be established.

Generally vague, the basic AI act is Taiwan's first stab at setting up a legal framework to regulate AI applications.

It contains 18 articles and will be available for public comment for 60 days until Sept. 13, after which the NSTC is expected to review it again and send it to the Cabinet for approval by the end of October.

If the Cabinet approves it, the law will then be submitted for deliberation and passage by the Legislative Yuan.

Under the proposal, the mitigation of risk is a broad theme, and it stipulates that there should be clear rules to reduce risk when using AI applications, such as the AI Risk Management Framework established by the U.S. National Institute of Standards and Technology.

Such a framework should mandate AI security standards and verification mechanisms along with accountability frameworks, the draft proposal said.

It also stipulates that personal data protection authorities assist regulators in avoiding the unnecessary collection, processing or use of personal data during the AI R&D and application processes.

Additionally, the proposal requires the government to regulate the use of AI by coming up with tools or methods that prevent the misuse of AI systems that would be harmful society or mislead users.

It says steps should be taken to prevent AI from causing damage to people's life, body, freedom or property, social order, or the environment, or from showing conflicts of interest, discrimination, bias, false advertising, and misleading or false information that would violate existing laws and regulations.

To protect workers' rights and interests, the government should provide employment guidance for those who become unemployed due to the use of AI, the draft act states.

Estimates have suggested that AI could cost about 30 percent of Taiwan's working population their jobs.