Will Depue, who worked on OpenAI's Sora, claims China has complete access to OpenAI and Anthropic's internal GitHub and Slack · Digg

/Tech1h ago

Will Depue, who worked on OpenAI's Sora, claims China has complete access to OpenAI and Anthropic's internal GitHub and Slack

Story Overview

An ex-OpenAI research engineer who contributed to Sora asserts that China holds unrestricted entry to OpenAI and Anthropic code repositories, chat channels, and internal notes, yet supplies no logs, screenshots, or incident details to back the statement.

1009601714278.3K

Original post

Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)@teortaxesTex#501inTech

If Chyna is so good at digital espionage, why wouldn't it be even better at architecture research? If you just argue that your colleagues are CCP agents, please let them go back already.

will depue@willdepue

there is no question, none at all, that china has full access to all of openai & anthropic’s github/slack/docs today

no disrespect to their independent research progress, but i wouldn’t be surprised if we see plausibly-deniable stolen arch methods in chinese oss models

4:51 PM · Jun 19, 2026 · 4.8K Views

Open Question

Basis for the claim remains thin

Depue points to private talks with security researchers plus his own time inside OpenAI, while acknowledging he sees no technical proof that would settle the matter.

Industry Shift

Access symmetry draws equal notice

Replies highlight that Western labs appear to hold similar visibility into leading Chinese models, raising the possibility that method borrowing flows in both directions without public attribution.

Sentiment

Some users praised claims that Western labs have full access to Chinese AI models while others dismissed allegations of Chinese access to OpenAI and Anthropic systems as unfalsifiable or McCarthyist.

Pos

50.0%

Neg

50.0%

5 comments with sentiment.

Cluster Engagement

Digg Deeper

No Digg Deeper questions have been answered for this story yet.

Related links

Governance Approaches to Securing Frontier AI

RAND.ORGVia

Why Security Comes First

SUBSTACK.COMVia

Posts from X

Most Activity

VIEWS5.8KLIKES202RETWEETS4REPLIES20

roon@tszzl

@willdepue Maybe they can figure out wth is going on here

will depue@willdepue

there is no question, none at all, that china has full access to all of openai & anthropic’s github/slack/docs today

no disrespect to their independent research progress, but i wouldn’t be surprised if we see plausibly-deniable stolen arch methods in chinese oss models

1h5.8K2023

BOOKMARKS9

kache@yacineMTB

A drama where a Chinese spy and an American CIA spook meet at openai while they're larping as OpenAI researchers and fall in love

will depue@willdepue

@CFGeek after talking at length with security researcher friends, and first-hand experience of openai security posture, i really can’t see this not being true

and even so, i don’t think anyone would dispute there being a handful of human intelligence sources per lab

1h4.2K609

elie@eliebakouch

there is no question, none at all, that all labs has full access to all of deepseek, kimi, glm, minimax tech report/model today

will depue@willdepue

there is no question, none at all, that china has full access to all of openai & anthropic’s github/slack/docs today

no disrespect to their independent research progress, but i wouldn’t be surprised if we see plausibly-deniable stolen arch methods in chinese oss models

1h5.1K1296

will depue@willdepue

@CFGeek after talking at length with security researcher friends, and first-hand experience of openai security posture, i really can’t see this not being true

and even so, i don’t think anyone would dispute there being a handful of human intelligence sources per lab

1h4.2K294

will depue@willdepue

i don’t know if this has happened yet, likely not, and they don’t seem to need the help much so far anyways. but it probably will

even just pointing at the papers to look hard at is often enough. having conviction, knowing what works is most of the work anyways

1h1.6K152

Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)@teortaxesTex

The really absurd thing is that Americans treat their AGI industry as an honest-to-goodness Manhattan Project to cybernuke the Chinese, but don't actually expect the latter to resist in any meaningful way except try to compete in products. The CCP are hippies, thus deserve death

flaw@flawedaxioms

It is really absurd that like 60% of the staff of US companies building potentially the most vital dual use tech ever are foreigners. This has to be making the admins head explode on the daily.

47m1.5K191

Miles Brundage@Miles_Brundage

I wouldn't put things in quiiiite so absolute of terms (there are at least some practical issues in exfiltrating everything given access IIUC) but broadly speaking, yeah it's a super serious/bad situation + I appreciate @willdepue calling attention to it

will depue@willdepue

there is no question, none at all, that china has full access to all of openai & anthropic’s github/slack/docs today

no disrespect to their independent research progress, but i wouldn’t be surprised if we see plausibly-deniable stolen arch methods in chinese oss models

58m1.7K111

Nathan Lambert@natolambert

@eliebakouch banger

elie@eliebakouch

there is no question, none at all, that all labs has full access to all of deepseek, kimi, glm, minimax tech report/model today

1h747130

kache@yacineMTB

Starring Adam Sandler and Jennifer Aniston

1h80712

shako@shakoistsLog

@tszzl @willdepue 100% chance you guys have hired spies

39m11271

Andrew Curran@AndrewCurran_

@teortaxesTex

will depue@willdepue

@CFGeek after talking at length with security researcher friends, and first-hand experience of openai security posture, i really can’t see this not being true

and even so, i don’t think anyone would dispute there being a handful of human intelligence sources per lab

1h1.7K41

Susan Zhang@suchenzang

just saving some new receipts here coincidentally on this particular thread

Susan Zhang@suchenzang

reminder to one day tell the tale from many moons ago of how a certain group within a certain lab pushed for extra surveillance/auditing of certain ethnicities (regardless of nationality)

傻逼透了

40m43611

Miles Brundage@Miles_Brundage

https://www.rand.org/pubs/research_reports/RRA4159-1.html

Miles Brundage@Miles_Brundage

I wouldn't put things in quiiiite so absolute of terms (there are at least some practical issues in exfiltrating everything given access IIUC) but broadly speaking, yeah it's a super serious/bad situation + I appreciate @willdepue calling attention to it

58m43001

Teortaxes▶️ (DeepSeek 推特🐋铁粉 2023 – ∞)@teortaxesTex

The real question: what Chinese lab can defend its original research from being stolen by Dario, now that he's built a SuperHaxx0R-1337 machine precisely to penetrate any security? Anthropic's bloated overpriced designs may suddenly get a big uplift…

1h3948

prinz@deredleritt3r

I find the most recent Chinese open-source models to be lacking in reasoning capabilities and generally sloppy for use cases that don't involve coding. In other words, distillation of coding abilities seems to be happening, and so is benchmaxxing, but overall these models are far, far away from the frontier.

This could be due to compute scarcity in China, and/or maybe they haven't quite infiltrated the U.S. labs to the extent you think.

1h8367

will depue@willdepue

and really, i think the slack/docs access is what really matters most. arch code is one thing, understanding theory of why or how something works is another

1h1.2K9

the tiny corp@__tinygrad__

@willdepue Why don't they just open source, cut out the middle man?

1h7615

will depue@willdepue

@CFGeek i love my chinese and israeli and russian and french and uk colleagues but odds are very high that at least one of my friends had a handler. a few once served in foreign intelligence units like hmmm

1h2819

Underground Thinker@Donogzs

@deredleritt3r @willdepue Does GLM 5.2 not make a difference?

1h624

ShitCockaSays@batcz

@willdepue > no disrespect

> unfalsifiable claim about entire country stealing

🤡

39m1293