South Korea has enacted its own national AI laws targeting trust and safety, the second major government after the EU to impose sweeping measures. But with the EU now pausing some of those restrictions under pressure from the Trump administration, South Korea could emerge as the global leader in AI safety regulation. The country has also ramped up development of its domestic models as it vies with Europe and the Gulf countries for the third top spot of global AI power, behind the US and China.
Passed in December 2024, the laws require human oversight on “high impact” AI, including technology that touches nuclear materials, education, drinking water, and medical devices. Products and services that use such “high impact” AI must disclose it, and generated media must be labeled as such to prevent being mistaken as human-made. The policy also outlines investment areas for the country, including incentives to attract foreign talent.
Penalties aren’t as severe as those in the EU rules. Someone who fails to label AI content in South Korea could face fines of up to 30 million won ($20,000), while the EU can impose fines of up to €15 million ($17.6 million) or up to 3% of a company’s annual revenue for a similar offense.