Overview
- The UN’s high commissioner Volker Türk said on Nov. 20 that roughly five companies hold more than 70% of global AI compute and called for international rules and independent oversight to prevent unchecked power over systems that affect rights and services.
- European institutions have already seen the danger: the European Commission detected an AI‑assisted cyberattack on Oct. 22 that led to temporary disconnection of some applications and CERT‑EU has warned that public AI vulnerability scanners such as ARTEX can speed automated attacks.
- Communication experts at a Barcelona conference said generative tools now produce content at massive scale and make messages more alike, which undermines source credibility and forces brands and journalists to compete for verified visibility.
- Business speakers and studies reported fast commercial uptake of agentic AI that automates tasks and spawns low‑cost competitors, offering big productivity gains for some economies but raising questions about job redesign, oversight and uneven benefits.
- Technology moves toward running models on devices promise lower latency and more data control but do not remove reliance on scarce chips or cloud updates, and recent export‑control disputes and detained advanced processors in Hong Kong show how supply chains and geopolitics can constrain access and security.