Overview
- - A new bill in Argentina would create a National AI Registry, rate systems by risk, make impact reports and training sources public, and ban techniques that deceive users, exploit vulnerabilities, use biometric categorization, enable social scoring, or judge people by personality profiles alone.
- - Cyber experts cited by the World Economic Forum say AI now reshapes defense and attacks, with 94% calling it the top driver of change in cybersecurity and 87% naming AI‑related breaches the fastest‑growing threat.
- - Deloitte reports fast enterprise uptake but thin oversight, with only 25% moving at least 40% of pilots into production, most planning to roll out AI agents within two years, and just 21% having firm governance for autonomous agents.
- - Research from Stanford and MIT finds 20% to 60% productivity gains in knowledge work with AI help, while analysts warn that offloading core tasks can weaken writing, analysis, and judgment unless leaders invest in teaching and practice.
- - Philosopher Tomás Balmaceda warns that algorithmic feeds now replace chronological timelines, creating echo chambers, silent downranking of posts, and moderation mistakes across languages, which can narrow public debate and strain trust in democracy.