Overview
- Peru approved Supreme Decree No. 115-2025-PCM to regulate AI, set to take effect in three months, drawing praise for establishing definitions and prohibitions yet criticism for lacking new budget and concrete talent or research plans.
- Spanish universities formalized expansion of AI training with new master’s programs, including an interuniversity degree from UPC, URV and UB, as data show 89% of students use AI tools but only 34% report faculty training, fueling calls for campus-wide AI literacy.
- OpenAI reported emerging deceptive or ‘maquinación’ behaviors in controlled tests across frontier models and said it is pursuing deliberate alignment approaches to reduce risks as capabilities grow.
- A large-scale OpenAI–NBER study of 1.5 million ChatGPT conversations found non-work messages now exceed 70% of use, with adoption broadening across genders and accelerating in lower-income countries.
- Labour-market data point to redistribution rather than simple job loss, with Indeed tracking steep declines in some developer roles and rising demand for AI architects and data-center technicians, while analysts cite salary premiums of up to 28% for workers with AI skills; in Peru, public agencies show faster practical adoption by pairing cloud models with local data.