Nvidia CEO Jensen Huang declares AGI has arrived following OpenAI's GPT-6 Astra release

Sep 4, 2026

Following OpenAI's September 3, 2026, release of its GPT-6 Astra model, Nvidia CEO Jensen Huang declared that artificial general intelligence (AGI) has officially arrived. Trained on over 100,000 Nvidia Grace Blackwell systems, Astra achieved historic benchmark scores, including 99.9% on ARC-AGI-3. While OpenAI President Greg Brockman welcomed the 'AGI era,' prominent skeptics like Gary Marcus challenged the claim, arguing that declaring victory without a consensus definition merely muddies the scientific waters.

Jensen Huang AGI declaration

▪Nvidia CEO Jensen Huang previously claimed during a March 23, 2026, Lex Fridman podcast appearance that artificial general intelligence had already been achieved.
▪Nvidia CEO Jensen Huang declared on September 6, 2026, that artificial general intelligence has arrived, following the release of OpenAI's GPT-6 Astra model.
▪During Nvidia's Q2 2026 earnings call on August 27, 2026, Jensen Huang stated that artificial general intelligence has already been achieved for many practical tasks.

OpenAI GPT-6 Astra release

▪OpenAI President Greg Brockman stated on September 3, 2026, that the release of GPT-6 Astra marks the beginning of the artificial general intelligence era.
▪OpenAI released its GPT-6 Astra model on September 3, 2026, describing it as the company's most advanced, intelligent, and ethically aligned system.

Nvidia Grace Blackwell infrastructure

▪Nvidia's data center business generated $89 billion in revenue for the quarter ending in August 2026, out of $96.2 billion in total quarterly revenue.
▪Nvidia CEO Jensen Huang announced that an additional 400,000 Nvidia graphics processing units are coming online in the near future.

GPU deployment expansion plans

▪Nvidia is developing an internal workforce model designed to operate over 400,000 AI agents alongside approximately 40,000 human employees.
▪OpenAI defines artificial general intelligence as highly autonomous systems that outperform humans at most economically valuable work.

AGI definition debate

▪AI pioneer Andrew Ng stated in 2026 that the technology industry remains decades away from achieving artificial general intelligence.
▪AI researcher Gary Marcus disputed Jensen Huang's declaration, arguing that the AI community lacks a consensus definition of artificial general intelligence.
▪OpenAI reported that GPT-6 Astra scored 98% on FrontierMath Tier 4, 99.9% on ARC-AGI-3, and 100% on ExploitBench.

Astra benchmark performance

▪The creators of the ARC Prize stated that GPT-6 Astra's benchmark success does not prove artificial general intelligence due to the tests' tightly bounded environments.
▪OpenAI stated that GPT-6 Astra met its internal critical cybersecurity threshold, demonstrating the ability to independently discover software vulnerabilities and plan exploits.

Debatable claims

▪GPT-6 Astra represents the arrival of artificial general intelligence
▪Artificial general intelligence should be defined by economic utility rather than human cognitive capacity