Photo via Washington Examiner

OpenAI defends firing researchers after they questioned company culture on third-party safety collaboration

Read the original at Washington Examiner ↗
Cite

Hallas, E. (2026, October 9). OpenAI defends firing researchers after they questioned company culture on third-party safety collaboration. Washington Examiner. https://www.washingtonexaminer.com/policy/technology/4761398/openai-defend-firing-researchers-safety-collaboration/

Dated when The Aporia collected the article, usually within hours of publication. Check the original for its exact date and byline.

Washington Examiner · collected 2026-10-09 · by Emily Hallas

Quick Summary

On Friday, OpenAI responded to allegations from three recently fired safety researchers—Tomek Korbak, Jasmine Wang, and Mikita Balesni—who claimed they were terminated for collaborating with external experts on AI safety issues. The company countered that the firings were due to policy violations regarding sensitive information rather than discussions about safety. OpenAI emphasized its commitment to third-party safety assessments, stating that it is working closely with independent organizations while also implementing internal procedures to manage such collaborations securely.
Written by AI on 2026-10-09, from this article's own text only (the story summary below draws on all the coverage).
See how they differ ↓

AI analysis

Story summary

On October 8, three former OpenAI security researchers—Mikita Balesni, Tomek Korbak, and Jasmine Wang—publicly accused the company of firing them for raising safety concerns about artificial intelligence rather than mishandling information as stated by OpenAI. The researchers fear their dismissals could stifle open dialogue within the organization regarding AI risks. OpenAI responded by stating it fired the trio due to a significant breach of trust involving sensitive information, contradicting the employees' claims that they were terminated for prioritizing safety over corporate interests. This incident highlights growing tensions in the tech industry between advancing AI capabilities and ensuring public safety, particularly amid recent security breaches involving AI systems.

Written for “OpenAI Safety Controversy” on 2026-11-20, grounded in this article and the 6 other(s) covering the same event.

How they differ

7 outlets covered “OpenAI Safety Controversy”.

What each outlet says

One AI summary per outlet of its latest article on this story, written from that article alone and checked against it. Open the article for the outlet's own words.

Washington ExaminerThis outlet9 Oct

On Friday, OpenAI responded to allegations from three recently fired safety researchers—Tomek Korbak, Jasmine Wang, and Mikita Balesni—who claimed they were terminated for collaborating with external experts on AI safety issues. The company countered that the firings were due to policy violations regarding sensitive information rather than discussions about safety. OpenAI emphasized its commitment to third-party…

OpenAI defends firing researchers after they questioned company culture on third-party safety collaboration (this article)
The Straits Times9 Oct

Three former OpenAI security researchers were fired on October 1 and accused the company of dismissing them for raising concerns about artificial intelligence risks. The researchers—Mikita Balesni, Tomek Korbak, and Jasmine Wang—claimed their termination is discouraging internal discussions on AI safety. In response, OpenAI stated that they violated company rules by mishandling sensitive information. This incident…

Fired researchers accuse OpenAI of ‘chilling’ safety efforts →
BBC9 Oct

Three former OpenAI researchers claim they were fired for raising safety concerns rather than mishandling information as stated by the company. On October 2nd, an OpenAI spokesperson told the BBC that the employees were dismissed due to a significant breach of trust involving sensitive data. The researchers—Mikita Balesni, Tomek Korbak, and Jasmine Wang—all formerly worked on AI safety at OpenAI and assert that…

Fired OpenAI researchers say they were let go for 'prioritising safety' →
Al Jazeera9 Oct

Three former OpenAI employees—Mikita Balesni, Tomek Korbak, and Jasmine Wang—claim they were fired for raising safety concerns about artificial intelligence. Firing allegations surfaced on Thursday following their dismissals last week. The ex-employees argue that OpenAI’s termination of them will discourage others from discussing AI risks openly. In response to the accusations, OpenAI stated that the firings…

Ex-OpenAI staff say they were fired for raising safety concerns →
CBS News9 Oct

OpenAI defended its decision to fire three AI safety researchers—Mikita Balesni, Jasmine Wang, and Tomek Korbak—for violating company policies on handling sensitive information. The company stated that the dismissals were not related to raising safety concerns but rather to breaches of trust beyond what was outlined in their public letter. OpenAI emphasized its commitment to maintaining a high degree of internal…

OpenAI defends firing AI safety researchers over "breach of trust" →
Global News9 Oct

OpenAI dismissed three safety researchers—Tomek Korbak, Jasmine Wang, and Mikita Balesni—for alleged policy violations involving sensitive information. The company claims it fired them due to a breach of trust, while the researchers contend they were terminated for prioritizing AI safety over corporate interests following an incident where OpenAI’s AI agents breached Hugging Face’s servers. The researchers fear this…

OpenAI fires 3 safety researchers in dispute over AI risks →
Show 1 more outlets
NPR9 Oct

Three former OpenAI employees, Mikita Balesni, Tomek Korbak, and Jasmine Wang, have accused the company of firing them over pretexts related to safety concerns or working with external researchers. This comes as OpenAI faces scrutiny for its handling of AI agent hacks and attempts to maintain a commitment to responsible technology development. OpenAI denies the allegations, stating that the employees were dismissed…

Fired OpenAI employees question the company's commitment to safety →

Where they agree and split

Framing

How the first headlines put it

Attributed to a source2
Stated as fact2
The headlines (4)

Fired OpenAI researchers say they were let go for 'prioritising safety' BBC · attributed to a source

Ex-OpenAI staff say they were fired for raising safety concerns Al Jazeera · attributed to a source

Fired researchers accuse OpenAI of ‘chilling’ safety efforts The Straits Times · stated as fact

OpenAI defends firing AI safety researchers over "breach of trust" CBS News · stated as fact

The full story: what happened, how it changed, and all 7 articles →

Compared with similar articles

This article reads leans left and hedges 30% of its claims. Each row says how that neighbour differs.
Global News
Hedges 1 of 14 claims Outlet record 63/100
CBS News
Hedges 3 of 24 claims Outlet record 67/100
The Straits Times
Hedges 3 of 19 claims Outlet record 64/100
NPR
Leans left Hedges 3 of 48 claims Outlet record 60/100
BBC News
Hedges 3 of 16 claims Outlet record 65/100
Al Jazeera
Hedges 3 of 30 claims Outlet record 60/100
Toronto Star
Leans left Hedges 1 of 2 claims Outlet record 64/100
“Article A describes a hearing where AI insiders warn NYC council about dangers of emerging technology, while Article B focuses on OpenAI defending firings of researchers who questioned company culture.”
Washington Examiner
Leans left Hedges 12 of 58 claims Outlet record 67/100
“The articles describe related developments but different specific events: one about a resignation and warnings from industry leaders, and another about OpenAI defending firings of safety researchers.”

Story

OpenAI Safety Controversy
Technology · 7 article(s) covering the same event. Compare every outlet's account →
How this article was measured signals, leaning charts, publisher record, byline, topics and claims

Signals How these are calculated →

Claims extracted
20
sentences that state something as happening or true
Uncertain
30%
6 of 20 hedged ("reportedly", "may", "according to")
Leaning
Leans left
of the writing, not the subject · beta estimate
Publisher signals
67/100
Washington Examiner's corrections and hedging, across everything collected from it. Not a measure of accuracy.
Outlets on this story
7
Technology
Same framing elsewhere
1
articles that frame the event the way this one does
Analyzed 2026-10-09 · how these are computed

How this is being covered How these are calculated →

Article leaning vs. publisher reliability
Source leaning vs. consistency

Publisher

Washington Examiner · 5208 article(s) · 11 correction(s) detected
Running correction rate · 11 correction(s)

Who wrote this

Emily Hallas
170 article(s) here · 1 carrying a prediction
A series of sudden outages across major sectors is triggering fear among U.S. shoppers, sending them to stores to prepare for what they believe could be an impending crisis.
Nearly 1 in 5 voters who elected President Donald Trump to a second term in 2024 believe the Democratic Party will flip the White House in the 2028 elections, according to a new poll.
From banks and aviation to the government and hotels, the past seven days have seen an unusual number of technological challenges, though experts are cautioning against attributing every spike in outages as being unprecedented, noting there could be “recency bias” at play.
Federal Reserve governor Lisa Cook on Tuesday revealed she will refuse to attend a committee hearing the White House spearheaded to investigate allegations that she committed mortgage fraud.
The inspector general’s office said the Pentagon could not provide sufficient evidence to the auditors “to support an opinion on the reliability of the financial statements” after 1,500 auditors examined nearly 30 agency components accounting for $4.8 trillion in assets.
A report this week that Iran could be plotting to target Washington, D.C., with a dirty bomb is stirring fears about how it could impact the city.
Vance was pressed on whether people have reason to be concerned about a possible larger impending threat, after all U.S. flights were grounded for hours last week, major banks, United Airlines, and Microsoft faced technology outages, and reports began spreading that Iran could be planning to target Washington, D.C., with a “dirty bomb.”
New York City Mayor Zohran Mamdani announced Monday plans to eliminate one of the city’s police department’s elite response units and impose significant restrictions on the force that law enforcement can use when responding to protests.
As dismal results trickled in for Republicans on election night last week, Trump first suggested he would be open to crafting legislation with Democrats on an immigration deal, and later suggested it could offer a path to legal status for “good people” who crossed the border illegally.
Allen, a 31-year-old California resident, was indicted in May on four charges, including attempting to kill the president at the April 25 gala, transportation of a firearm and ammunition in interstate commerce with intent to commit a felony, and use of a firearm in a crime of violence.
Also by Emily Hallas
Nothing else under this byline is closely related to this article, so these are simply their most recent.
All 170 articles by Emily Hallas →

Topics

Anthropic METR OpenAI Washington

Subjects

OpenAI ORG · 10× Anthropic ORG · 2× Balesni PERSON · 1× Jasmine Wang PERSON · 1× Korbak PERSON · 1× METR ORG · 1× Mikita Balesni PERSON · 1× Sam Altman’s PERSON · 1× Tomek Korbak PERSON · 1× Wang PERSON · 1×

Narrative

In a public statement, the company said its investigation found the trio “violated clear policies on handling sensitive information,” and uncovered “a significant breach of trust beyond what’s outlined in the letter they published.”
framing: mixed · carried by 1 article(s) · first seen 2026-10-09
Sam Altman’s OpenAI on Friday pushed back against allegations by three fired employees that the company could be undermining external checks on artificial intelligence, stating that the AI firm views safety and research debates as “highly critical” to success.

Claims (20 extracted, 6 hedged)

Sam Altman’s OpenAI on Friday pushed back against allegations by three fired employees that the company could be undermining external checks on artificial intelligence, stating that the AI firm views safety and research debates as “highly critical” to success. uncertain
firm → push → success
The saga began on Thursday, when three safety researchers at OpenAI released a letter announcing their termination by the company. asserted
researchers → begin → company
Tomek Korbak, Jasmine Wang, and Mikita Balesni suggested they were fired for collaborating with outside experts on safety concerns, intimating that OpenAI is becoming less open to its researchers working in tandem with independent entities to create safety mechanisms for AI. OpenAI fired back the following day. uncertain
OpenAI → suggest → AI
In a public statement, the company said its investigation found the trio “violated clear policies on handling sensitive information,” and uncovered “a significant breach of trust beyond what’s outlined in the letter they published.” asserted
they → say → letter
The statement offered little detail on why the three were fired, but claimed: “These decisions were not about them raising safety concerns.” asserted
them → offer → concerns
“We are actively finalizing contracts with third-party safety assessors and will announce details in the coming weeks. asserted
We → finalize → weeks
People across the company have been working really hard on getting these partnerships up and running,” OpenAI’s statement read. asserted
statement → work → company
“We are committed to embedding external assessors and continue to make close collaboration with independent safety organizations a core part of our safety work. asserted
collaboration → commit → work
Many of our researchers already work with 3p safety organizations productively.” asserted
Many → work → organizations
In their letter, Korbak, Wang, and Balesni said OpenAI has historically been open to safety collaboration with third parties, writing that traditionally, employees “could raise safety concerns and disagree openly, and were encouraged to draw on the expertise of independent safety organizations.” uncertain
employees → say → organizations
The trio suggested that culture is changing, warning “we do not believe the path to superintelligence can be navigated safely if the people closest to the risks can no longer work in high-trust, high-bandwidth ways with each other and with third parties.” uncertain
people → suggest → parties
“We are concerned that our firings may be used to justify ending OpenAI’s work with METR, or otherwise providing external auditors much more limited access and scope,” their letter read. uncertain
letter → concern → access
“Those of us who work on safety see risks before anyone else, and we rely on close collaboration with outside experts to work out how to address them. asserted
we → work → them
The freedom to do so without fear, and to have well-defined internal procedures that enable this work, is itself an essential safety mechanism.… asserted
that → have → work
It is that culture we are writing to defend.” asserted
we → write → ?
OpenAI has called for more safety controls and third-party oversight in recent weeks, after another of its former employees sparked the recent wave of national panic about AI, triggering a push in Washington to reform the industry. asserted
another → call → industry
“I spent the last three years doing pretraining research at both OpenAI and Anthropic. asserted
I → spend → OpenAI
Neither company is acting responsibly,” former Anthropic and OpenAI researcher Jacob Coxon said in a post to X. “The people building AI earnestly believe that it could kill us all by the end of the decade. uncertain
it → act → decade
If anything, many executives and senior researchers will couch their phrasing in the press to sound sensible – but I hear the same people express fear privately. asserted
people → couch → fear
No other human activity poses this level of danger.” asserted
activity → pose → danger
Give feedback · no sign-in needed

No sign-in needed for now. Your note is sent without a name, so it cannot be answered.

History Support