KADIN Indonesia official calls for AI safety testing in users' languages

KADIN Indonesia official Firli Ganinduto said artificial intelligence should be tested for safety in the language its users will speak and within their cultural context, RIA Novosti reported from a forum in Moscow.

Ganinduto heads the Indonesian chamber of commerce's digital ecosystem agency. He spoke at a session examining AI development for languages with limited collections of digital text.

Safety controls and restrictions receive the most intensive testing in languages with abundant digital resources, particularly English, he said. That can leave uneven protection when systems are used in other languages.

Ganinduto argued that successful English-language tests should not automatically be treated as a guarantee of equivalent accuracy or safety in languages with less data. He called for evaluations reflecting how the systems will actually be used.

The second Global Digital Forum took place in Moscow from October 8 to 10. The report of his remarks does not present results from a specific comparative model test.