OpenAI's Friar, Astra & Navier-Stokes

OpenAI has set out its case for why growing consumer and enterprise use of its models should compound into a larger business, in a blog post from chief financial officer Sarah Friar published on 8 September that leaned heavily on the recent launch of GPT-6 Astra and, in an aside, a claimed proof of a decades-old mathematics problem.

Friar said OpenAI's products now reach more than one billion weekly active users and 2.5 million businesses, and argued that research advances feed directly into ChatGPT, ChatGPT Work, Codex and API-based applications, giving a single model improvement multiple routes to revenue. She cited internal data showing that people on individual ChatGPT plans send roughly 50% more daily messages six months after signing up than in their first month, and have tried around twice as many distinct tasks by that point. Within OpenAI's own research organisation, she said, agents now contribute the equivalent of 3.1 agent-workdays of effort for every workday of human labour, based on an August measurement using an eight-hour working day as the baseline.

Much of the argument rests on GPT-6 Astra, which OpenAI has described elsewhere as its most capable and best-aligned model to date, now available in ChatGPT Work, Codex and the API. OpenAI says Astra leads on Terminal-Bench 4.0 at 57.9% accuracy, ahead of its predecessor GPT-5.6 Sol and Claude Fable 5.1, while costing less per task, and that it produced unintended outcomes in sensitive business scenarios 89% less often than GPT-5.6 Sol during internal safety testing. The company also pointed to Jalapeño, its first custom inference chip, which it says delivered 1.5 to 1.9 times the peak token throughput per watt of commercial systems tested, with deployment planned by year-end.

Friar's post also referenced, briefly, a separate announcement that an internal OpenAI model — more capable than GPT-6 Astra and still in training — had produced a proof that the Navier-Stokes equations, which describe fluid motion and have stood as an open problem for around 90 years, can develop a singularity in finite time. OpenAI said the result, reached using a coordinated system of roughly 10,000 concurrent agents over 88 hours, was intended to illustrate the pace of AI progress rather than to claim the associated Millennium Prize. The company said it had no visibility into related work by Tristan Buckmaster, a New York University mathematics professor, on a concurrent Euler-equations problem until he made it public.

Sarah Friar, OpenAI's chief financial officer, said the combination of broad product reach and a "full stack" compute strategy gives the company "conviction" in its ability to lead through successive generations of AI, funded by revenue from growing usage.

Swiss Cheese Defences- Identity fraud goes industrial and off-the-shelf

Ofer distinguishes between two current attack patterns: highly sophisticated, professionally coordinated deepfake and injection attacks designed to beat detection outright, and a much larger volume of lower-effort, high-scale attempts that rely on bombarding systems rather than disguising themselves well. He argues the real story right now is industrialisation of scale rather than uniform improvement in quality, though both are accelerating in parallel. The conversation covers why agentic AI is opening a new front in identity fraud, particularly the unresolved problem of tying an AI agent’s identity back to the human who deployed it and the permissions it holds. Ofer is candid about the current state of agent defences, describing them as underdeveloped and easy to hijack or poison, comparing the current state of play to Swiss cheese. He also discusses cross-industry fraud detection, including how signals, rather than raw data, are now being shared across platforms including AU10TIX and Reality Defender to surface fraud rings invisible to any single organisation. The conversation also covers explainability as one of AI’s most underdeveloped capabilities, with Ofer arguing that flagging a session as fraudulent without a credible, defensible reason will increasingly run into regulatory and practical limits. Elsewhere, the discussion covers credential laundering and the manufactured construction of fake digital histories and footprints, and the shift toward digital ID wallets that Ofer believes will make physical document fraud increasingly rare. He closes on what he calls “agentic avatars,” AI systems capable of holding a full visual and verbal conversation on someone’s behalf, and why he expects identity verification to have to become genuinely immersive across every form of media as a result.

BrightTALK