Overview
- Three in four executives now view AI agents as collaborators, with 35% of organizations exploring agentic systems and another 44% planning deployments, according to the latest MIT Sloan–BCG survey.
- OpenAI’s ChatGPT has rolled out Atlas, an Agent Mode that retains browsing history, and instant payments for Shopify and Etsy, drawing criticism over blurred lines between recommendations and paid promotion and prompting some feature removals in December.
- China’s cyber regulator released draft rules for AI that simulates human personalities, requiring usage warnings, emotional state monitoring, and mandatory intervention when addiction or extreme emotions are detected, alongside stricter lifecycle security controls.
- Worker wellbeing is under strain as studies link ‘AI anxiety’ to lower performance, while separate research shows people struggle to spot AI‑generated faces—training helps but still yields only 64% accuracy for super‑recognizers—bolstering UNAM’s call for ethical frameworks and caution on high error rates in AI/plagiarism detectors.
- Market analysts flag more volatile trading ahead despite clearer monetization for infrastructure and cloud providers, urging selectivity as competition and valuation pressures build into 2026.