Daily AI Brief: November 16, 2026

The EU's deepfake ban lands in 16 days, OpenAI's AI agents can now spend your money, and India just announced its first government AI university.

The EU deepfake ban arrives December 2 — 16 days away

The EU's prohibition on AI-generated non-consensual sexual deepfakes and AI-produced child sexual abuse material takes effect on December 2, 2026. Transparency rules requiring platforms to label AI-generated content — including synthetic voices, images, and video — have been in force since August and face a final compliance deadline on the same date.

Why it matters: This is the most sweeping deepfake enforcement deadline in the world so far. Platforms serving EU users are scrambling to build takedown pipelines in days. India is watching closely — MeitY's own consultation paper on deepfakes and AI safety was expected by this month and has not yet appeared.

ChatGPT agents can now spend your money — if you're a US Pro subscriber

OpenAI's ChatGPT Wallet, launched November 13, lets AI agents make purchases on your behalf. You enter your card once, then approve each transaction individually. It is currently limited to US Pro subscribers aged 18 and over. The vision: an AI assistant that doesn't just recommend things, it buys them.

Why it matters: Handing an AI agent real purchasing power is a significant line to cross. It shifts AI from a tool that tells you what to do into one that acts on your behalf in the real world — with real money. Watch how quickly this expands beyond the US.

OpenAI, Anthropic, Microsoft sign AI safety charter for minors at Paris Peace Forum

On November 11, four AI companies — OpenAI, Anthropic, Microsoft, and Black Forest Labs — signed a non-binding charter at the Paris Peace Forum committing to stronger AI safety and privacy protections for children and teenagers. The charter does not carry legal force, but signals that the industry is trying to self-regulate ahead of government mandates.

Why it matters: Parents and educators in India should know that the world's leading AI labs are starting to draw child-safety lines themselves. Non-binding today; likely to become binding standards tomorrow as regulators take note.

India's first government AI university is coming to Bengaluru

Karnataka is establishing India's first public AI university in Bengaluru, with certificate, diploma, and reskilling programmes launching as early as January 2027. Separately, the state has rolled out a specialised AI curriculum for Class 9 students — covering statistics, calculus, linear algebra, and probability through AI-focused modules — across 1,800 government high schools.

Why it matters: India is building AI education infrastructure at scale before its national AI law even exists. If you are a student or teacher in Karnataka, the window to get trained early is now. If you are elsewhere in India, your state is likely to follow.

AMD's new AI chips target science and high-memory workloads

AMD expanded its Instinct MI400 GPU series with two new chips: the MI430X, designed for scientific computing, and the MI450X, built for high-memory AI workloads. The launch deepens AMD's challenge to NVIDIA in the market for AI training and research hardware, where Indian universities and startups are increasingly buying compute.

Why it matters: More competition in AI chips means more options and potentially lower costs for India's growing AI infrastructure push. The IndiaAI Mission's 38,000+ GPUs were heavily NVIDIA — alternatives are worth watching.

California now restricts AI in hiring decisions — first US state to enforce it

California's Civil Rights Department established rules restricting how employers can use AI tools in hiring, promotion, and employment decisions. Automated systems that screen resumes or score interviews must now meet anti-discrimination standards, and employers can be held accountable for biased AI outputs even if they didn't build the tool.

Why it matters: If you have ever wondered whether an AI rejected your job application, California is the first place trying to hold employers legally responsible for that. Indian employment regulators and IT hiring companies are watching.

India's national AI rules are overdue — and the window is closing

India's Ministry of Electronics and IT (MeitY) was expected to publish a consultation paper on AI safety, deepfakes, and startup compliance by November 2026. As of today, it has not appeared. The EU's December 2 deadline is bearing down. The US is enforcing. And Indian states — Karnataka, Goa, Telangana — are writing their own rules rather than waiting. The longer the Centre waits, the more fragmented India's AI governance will become. This is worth watching every day this month.

FAA deploys AI for air traffic + Malaysia moves on data center oversight

The US Federal Aviation Administration has expanded SMART — an AI tool advising air traffic controllers on congestion management — across the entire country. Meanwhile, Malaysia is drafting rules to monitor power use at AI data centers and crack down on smuggling of restricted AI hardware, an early sign that GPU supply chains are becoming a national policy issue across Asia.

All daily AI briefs