Episode overview

OpenAI CFO Sarah Friar & Reid Hoffman on where AI goes from here | Masters of Scale Summit 2026

Masters of Scale · 26m · 2 Egleze moments

What happened in this episode?

OpenAI CFO Sarah Friar tells Reid Hoffman that OpenAI is seeing rapid shifts from “chat” use of ChatGPT to “agent” and “do” use cases, and says 1.2 billion people use ChatGPT weekly. She points to commercial scaling, including a claim that OpenAI has 100,000 advertisers operating in 89 countries, arguing that ad revenue helps keep consumer access free in line with the company’s mission.

On safety and security, Friar says OpenAI recently “held back” its most powerful model under a “pacing the frontier” approach, claiming it solved the Navier–Stokes millennium maths problem and advanced hundreds of other theorems, while emphasising time to assess safety and alignment. She also describes cyber risks from powerful models and says OpenAI worked quietly with the PHP project after finding vulnerabilities that she claims could have affected “about 70% of websites”, aiming to patch them before attackers found them.

Friar frames trust-building through applied outcomes, highlighting healthcare usage and claiming enterprise pilots that reduce paperwork and shorten hospital stays. She describes uneven adoption between “frontier” and average firms, and says OpenAI is expanding deployment options, including being “live on GCP” alongside other major cloud providers. She also describes engagement with governments, including meetings in Dublin during Ireland’s EU presidency, and discusses community relations around data centres, citing an $80m grant tied to a planned Georgia site and funding for free school lunches as an example of local consultation.

What are the key points?

Friar claims 1.2 billion people use ChatGPT each week and argues keeping access free is central to OpenAI’s mission.
She says OpenAI has “held back” its most powerful model for safety reasons, and claims it solved the Navier–Stokes millennium maths problem.
Friar says OpenAI worked with the PHP project to patch vulnerabilities she claims could have affected “about 70% of websites”.
She cites healthcare as a trust-building wedge, including claims about heavy consumer use for health questions and pilots to reduce paperwork and hospital time.
Friar says “frontier firms” use far more agentic tokens than average firms, and predicts fast catch-up across industry.
She says OpenAI is now “live on GCP”, adding to deployments across other major cloud providers.
Friar describes community engagement around data centres, including an $80m grant for a Georgia community and funding free school lunches.

Who is speaking?

Named speakers identified in Egleze's source-linked moments from this episode.

Sarah Friar

Which topics are covered?

AI & Tech Crime & Justice Science

How is this page sourced?

Programme: Masters of Scale
Duration: 26m
Original episode: Open the original source ↗

Egleze verifies the speaker, source and timestamp for each linked moment. Unless a moment is explicitly marked independently verified, statements remain attributed claims by the named speaker rather than established facts.

Watch original episode More from Masters of Scale

2 moments from this episode

Source-linked · editorially selected

AI & Tech

OpenAI CFO Sarah Friar says it helped PHP fix security flaws

Sarah Friar · at 3:54 · Attributed speaker claim

This morning the journal just ran an article about PHP where we've been working with them since the end of the summer when we discovered using actually a N minus one model 56 a set of vulnerabilities in PHP which really means the internet. So about 70% of websites would have been impacted would have brought down probably a lot of the internet. So we've been working in a very secret way to get those vulnerabilities fixed before the offense discovers them.

Sarah Friar, OpenAI’s CFO, says OpenAI has worked with the PHP project since late summer after finding vulnerabilities using an earlier model. She says the flaws could have affected about 70% of websites and were fixed before attackers learnt of them.

Read this moment → Check original at 3:54 ↗

AI & Tech

OpenAI CFO Sarah Friar says withheld model solved Navier–Stokes problem

Sarah Friar · at 5:50 · Attributed speaker claim

Right now we have paced the frontier most recently by actually holding back our most powerful model partially because we want to make sure it's the safety and alignment is at a place where our researchers have had time to really think it through. That model is the model that solved the Navier Stokes Millennium Math problem and also about 350 more theorems have certainly moved along.

Sarah Friar says OpenAI has held back its “most powerful model” while it reviews safety and alignment. She also claims the withheld model solved the Navier–Stokes Millennium Prize Problem and advanced about 350 other theorems, without offering verification details in the clip.

Read this moment → Check original at 5:50 ↗