OpenAI’s latest GPT-6 Astra release in early September 2026 prompted executives including Greg Brockman to describe the model as potentially marking entry into the AGI era, with strong benchmark results in agentic coding, cybersecurity, and long-horizon tasks. Traders still assign 73% probability that the company will not formally announce AGI achievement before 2027 because leadership continues to treat the term as poorly defined and has avoided a definitive internal or public declaration. Internal roadmaps target an automated AI researcher only by March 2028, while recent safety pauses and alignment concerns underscore caution around premature claims. The market-implied odds reflect skepticism that competitive pressure or capability jumps will overcome these definitional and safety hurdles in the remaining months of 2026.