Group 1: AI and Copyright Developments - Meta has stated that it does not currently use unpublished user photos to train its AI models, despite previous reports suggesting otherwise [2][3] - Recent U.S. court rulings have determined that using published works to train AI models can fall under "fair use," with two cases involving Anthropic and Meta providing legal clarity [4][5] - The rulings indicate that while AI companies may have some leeway, the methods of data collection must still be scrutinized to avoid infringement [4] Group 2: Regulatory Changes - The revised Anti-Unfair Competition Law in China will take effect on October 15, 2025, aiming to curb "involution" in competition and establish fair competition review systems [5][6] - The new law prohibits large enterprises from abusing their dominant positions to delay payments to small and medium-sized enterprises [6] - The law also restricts platform operators from forcing other businesses to comply with pricing rules that disrupt market order [6] Group 3: Personal Information Protection - The National Cybersecurity Center reported that 45 mobile applications were found to be illegally collecting and using personal information without user consent [7] - In Shanghai, authorities are actively addressing the misuse of AI technologies, particularly focusing on the protection of personal information rights [8] - Ongoing enforcement actions aim to combat AI misuse, including the generation of inappropriate content and violations of personal data rights [8]
Meta否认用相册图训练AI|南财合规周报(第196期)