AI Misuse Countermeasures Expand: Deepfakes, Scraping, and Workplace Surveillance Regulations Spread

AI 안전 | Sat Jul 18 2026 00:00:00 GMT+0000 (Coordinated Universal Time) | 5 sources

AI misuse prevention measures expand with TikTok's AI deepfake detection, Patreon's AI bot blocking, and San Francisco's demand to remove nudify apps.

Analysis

[TikTok] began testing AI likeness detection tools ^[1]

Opt-in test targeting select U.S. creators
Real-time selfie and ID identity verification via Jumio
Scanning and reporting features for unauthorized AI deepfake content

[Patreon] introduced active blocking policy against AI training bots ^[2]

Leveraging Cloudflare's AI Crawl Control technology
Direct access blocking instead of robots.txt requests
Weekly access attempts dropped from thousands to zero in testing

[San Francisco City Attorney] demanded Apple and Google remove AI nudify apps ^[5]

Cease-and-desist notices targeting 13 face-swapping apps
Demand to stop facilitating non-consensual deepfake image generation
California law prohibits supporting deepfake porn services

[OpenAI] unveiled GPT-Red automated red-teaming system ^[3]

Automating software system safety evaluations
Replacing traditional human tester team approach
Automated exploration of system vulnerability discovery methods

[Kaiser Permanente] faced controversy over AI workplace surveillance degrading nurse work and patient care ^[4]

Performance reviews triggered when calls exceed 15 minutes
AI evaluates nurses' empathy and tone of voice
Key issue in California Nurses Association contract negotiations

AI Misuse Countermeasures Expand: Deepfakes, Scraping, and Workplace Surveillance Regulations Spread

Analysis

Sources