Overview
- Three in four executives now view autonomous AI agents as collaborators, with 35% exploring and another 44% planning deployments, according to the MIT Sloan–BCG survey.
- OpenAI’s ChatGPT Atlas, Agent mode and instant payments fold shopping into conversations, drawing user pushback that led to partial rollbacks in December and raising concerns about native advertising and disclosure.
- China’s cyber regulator proposed rules for services that simulate human personalities, requiring mood monitoring, interventions for addiction or extreme emotions, lifecycle security controls and strict content limits.
- UNAM researchers urged ethical governance and pedagogical frameworks for AI in education, warning that automated plagiarism and AI-detection tools can err by 30%–70% and stressing critical literacy over bans.
- Analysts report a more cautious stance on AI valuations as competition and return expectations rise, with guidance to be selective and favor firms with tangible exposure in cloud, chips and data-center infrastructure.