Bloomberg reports that the race among leading AI labs is increasingly being organized around artificial general intelligence, or AGI — a term that has become central to research plans, marketing and policy arguments even though companies have not agreed on what it would actually mean in practice. The report says tech companies in the US and China have spent hundreds of billions of dollars on data centers, chips and talent in the nearly four years since ChatGPT’s release. The initial objective was to build more capable AI models; increasingly, Bloomberg says, the stated goal is a broader and more powerful capability described as AGI. That shift matters because AGI is no longer only a speculative label in AI circles. According to Bloomberg, it has become a rallying point for leading labs and a reference point in policy debates, including concerns among some US lawmakers and AI leaders about the national security implications if another country — with China specifically cited in the report — reaches the milestone first. OpenAI is now putting that language closer to its product roadmap. Bloomberg reports that in early September, OpenAI began rolling out a more powerful AI model called GPT-6 Astra. The company said the model might one day be viewed as the early stages of artificial general intelligence. OpenAI President Greg Brockman underscored that framing at a press briefing for the model. “Welcome to the AGI era,” he said, according to Bloomberg. The unresolved issue is definitional. Bloomberg’s account says AI companies are racing toward AGI while still lacking agreement on what AGI would look like or how it might be achieved. That makes the term both strategically important and difficult to evaluate: it can guide investment and policy while remaining hard to measure against a shared standard. Who benefits: Leading AI labs with access to capital, compute and scarce technical talent benefit most from an AGI-centered race. OpenAI also gains a stronger narrative around GPT-6 Astra by tying the rollout to the possibility of early AGI capabilities, as Bloomberg reports. Who's exposed: Companies and policymakers are exposed to ambiguity: AGI is being used in strategic and national-security debates despite lacking a settled definition. Buyers and investors should be careful not to treat AGI claims as directly comparable unless labs define the benchmark clearly.