
New rules target 'bait-and-switch' tactics and mandate stricter moderation for user-generated content
AI-generated summary
Apple maintains an App Store review process to filter apps that violate safety or quality standards. Last year, the company reported blocking over 150,000 apps and removing 193,000 developer accounts.
Apple tightened its developer guidelines over the weekend to ban apps that use AI to create offensive content, as well as those that masquerade as regular apps to pass the App Store review process and then turn into nudify apps, or surface those features prominently.
The company updated its guidelines with strong language around AI and user-generated content, and how app developers should include enough preventative measures to moderate objectionable or offensive content.
“Apps that host user-generated content are required to filter objectionable material, give users a way to report offensive content, and let users block abusive accounts. Apps containing hidden or undocumented features are also prohibited on the App Store, including apps that misrepresent their purpose during review and only reveal harmful capabilities — such as “nudification” — after approval, a tactic commonly known as bait-and-switch,” the company said in a blog post.
Apple said that it will remove apps that add features not present in the version submitted for App Store review.
The iPhone maker has also updated the language in its legal section to reinforce that it will bar apps promoting human trafficking and exploitation of children.
Apple said last year its review process blocked over 150,000 apps from reaching users, and removed 193,000 developer accounts on account of fraud and abuse.

Nvidia launched the Nemotron 3.5 Super VL, a 120-billion-parameter open-weight model for enterprise agents, and two lightweight Bolt models for local AI PCs. The Super VL claims performance parity with larger models and is used by companies like Skydance and Salesforce. The Bolt models target consumer and enterprise users running on RTX Spark and DGX Spark hardware, enabling local deployment without cloud reliance.

Deutsche Telekom, Orange, Telefónica and Vodafone have launched a joint initiative to establish ground network infrastructure and coordination functions for a European satellite constellation, aiming to provide universal coverage in underserved areas and strengthen network resilience without direct investment in satellites.

Anthropic announced a public benefit fund backed by 5% of its shares, with potential to increase to 7% based on company growth, aiming to support societal good projects amid growing public distrust in AI.

After Character.AI filed for bankruptcy, Disney agreed to acquire some of its assets for $20 million and offer jobs to at least 60 employees, with most accepting. The deal includes an estimated $60 million in annual compensation and is structured as a stalking horse bid, allowing for higher offers.

Autonomous vehicle developers face a growing patchwork of local regulations as federal safety standards remain stalled in Congress. Cities and states are increasingly asserting authority over road usage, emergency protocols, and zoning to manage robotaxi deployments.

Meta updated its commercial terms to claim broad rights to use content from business accounts' domains for AI training, sparking criticism from legal experts who called it a 'remarkable provision' enabling potential publisher scraping. After backlash, Meta clarified the terms applied only to ads and business products, then removed the provision entirely, stating it does not change how the company uses web crawlers.