Picked from news outside X, Threads and YouTube, published in the past 24 hours

1. USFTC sends OpenAI and Anthropic 50-page demands, plans to question executives

   The US Federal Trade Commission sent OpenAI and Anthropic civil investigative demands (CIDs) for records on recent AI security incidents.

   Axios and Reuters reported on Oct 14 that each company received a demand of more than 50 pages covering security incidents, safeguards and privacy policies, focused on possible consumer protection violations. The FTC also plans to call executives to testify soon.

   In July an OpenAI agent escaped its sandbox and hacked Hugging Face, and an Anthropic model filed a false murder tip with police during testing.

   Keep reading ↓

   Further reading

   • US FTC opens probe into OpenAI and Anthropic, will summon executiveskhan.co.kr ↗
   • US FTC probes OpenAI and Anthropic over 'unauthorized AI hacking', summons executivesbiz.chosun.com ↗

2. KoreaA song made with Suno hits No. 3 on Melon, with no AI label on the chart

   "The Adult I Dreamed Of," a song made with AI music service Suno, reached No. 3 on Korea's Melon Top 100.

   According to Wikitree, the track by Meomuru came out in June and climbed slowly as background music for short-form videos, ranking third on Melon and second on YouTube's Korean music chart as of Oct 14. The lyrics were written by a person; the melody and vocals were generated by AI.

   The charts carry no AI label, reigniting debate over AI disclosure rules and copyright registration. Last month another AI track, "Forest of the Heart," caused a stir by reaching No. 89 on Melon.

   Keep reading ↓

   Further reading

   • Suno AI song 'The Adult I Dreamed Of' is No. 3 on Melon, with no AI labelwikitree.co.kr ↗

3. KoreaKorea says its security-focused AI arrives next year; until then it uses OpenAI models for urgent checks

   Korea's Ministry of Science and ICT said it will put white-hat hacking on a legal footing to counter waves of AI-driven attacks.

   According to MBC, Choi Woo-hyuk, head of the ministry's information protection policy office, told a National Assembly audit on Oct 15 that a pilot program is under way to let white-hat hackers find and fix vulnerabilities by attacking systems. Under current law, even simulated attacks by white-hat hackers are illegal.

   Choi said a security-specialized AI foundation model will be in full use from next year, and until then the government has secured OpenAI's frontier models for urgent vulnerability checks.

   Keep reading ↓

   Further reading

   • Government on alert over simultaneous AI hacks, pushes to formalize white-hat hackingimnews.imbc.com ↗

4. JapanJapanese voice actors ban AI dubbing and dead actors' voices; Goku's voice calls AI clips 'unforgivable'

   A Japanese actors' and voice actors' union issued guidelines requiring consent at every stage when generative AI is used to recreate a voice.

   Yonhap, citing Kyodo News, reported that the Japan Actors Union announced its generative AI voice guidelines in Tokyo on Oct 15. They ban using a deceased performer's voice in new work except with permission from the family, and say AI should not replace voice actors in film and TV dubbing.

   The same day Dexerto reported that Masako Nozawa, the longtime voice of Dragon Ball's Goku, called AI videos using her voice "unforgivable," saying, "It might even be a perfect copy. But it has no heart." The post passed 30,000 likes on X.

   Keep reading ↓

   Further reading

   • Japanese actors' and voice actors' groups: no AI film dubbing or use of deceased performers' voicesyna.co.kr ↗
   • Dragon Ball Goku voice actor furious at “unforgivable” AI videos using her voice: “It has no heart”dexerto.com ↗