According to Anthropic (@AnthropicAI), the SGTM (Stochastic Gradient Targeted Masking) unlearning method is significantly more resilient than previous approaches. Specifically, it requires seven times more fine-tuning steps to recover forgotten knowledge using SGTM compared to the RMU (Random Masking Unlearning) method. This finding highlights a critical advancement for AI model safety and confidential data retention, as SGTM makes it much harder to reintroduce sensitive or unwanted knowledge once it has been unlearned. For enterprises and developers, this strengthens compliance and data privacy opportunities, making SGTM a promising tool for robust AI regulation and long-term security (source: Anthropic, Twitter, Dec 9, 2025).

According to Sawyer Merritt, Tesla has introduced a new prompt in its Robotaxi app requesting user consent for enhanced in-cabin data sharing. This opt-in feature specifically targets two critical areas: improving the vehicle's autonomous driving capabilities and supporting customer service features. Tesla clarified that analytics are anonymous and in-cabin data is only linked to users in the event of a safety incident or support request (source: x.com/Tesla_App_iOS/status/1993160819715027322). This move reflects Tesla's continued focus on using real-world data to refine its AI algorithms for autonomous vehicles, opening new opportunities for AI-driven mobility solutions and data-driven vehicle safety enhancements.