Photo via Mother Jones

OpenAI’s “Independent Assessment” Doesn’t Look Independent at All

Read the original at Mother Jones ↗
Cite

Nguyen, A. (2026, November 17). OpenAI’s “Independent Assessment” Doesn’t Look Independent at All. Mother Jones. https://www.motherjones.com/politics/2026/11/openai-altman-independent-assessment-oversight-control-ai-tech-security/

Dated from the outlet's own feed. Check the original for its byline.

Mother Jones · collected 2026-11-17 · by Alex Nguyen

Quick Summary

OpenAI announced on Monday that four independent assessors will evaluate the safety and effectiveness of its AI models. However, critics argue that the assessors lack full access to OpenAI’s technology and that the company retains control over which staff members the assessors can interact with. Previous assessments faced limitations, such as short on-site investigation times and reliance on potentially unreliable AI agents for analysis. This new round of assessments includes labs like METR, Apollo Research, Transluce, and Goodfire, but their access is still subject to OpenAI’s approval and restrictions.
Written by AI on 2026-11-17, from this article's own text only (the story summary below draws on all the coverage).

AI analysis

Story summary

OpenAI announced on Monday that four independent assessors will evaluate its internal AI models and test the effectiveness of its safety features, amid ongoing security issues and declining public trust in its technology. However, critics argue that the assessors, including METR, Apollo Research, Transluce, and Goodfire, have limited access to OpenAI’s systems and that OpenAI retains control over which staff members the evaluators can interact with. This limited access contradicts CEO Sam Altman's September commitment to provide assessors with "employee-like access." Previously, in July, METR staff and a researcher from another AI safety organization conducted an investigation for two days, later extended to a total of six days, raising concerns about the thoroughness and independence of the assessment process.

Written for “OpenAI Assessment Criticism” on 2026-11-18, grounded in this article and the 0 other(s) covering the same event.

Compared with similar articles

This article reads leans strongly left and hedges 10% of its claims. Each row says how that neighbour differs.
New York Post
Hedges 4 of 23 claims Outlet record 66/100
“The articles describe different events: the lawsuit filed by California against OpenAI versus OpenAI's announcement of an internal assessment of its AI technology.”
The Guardian
Leans strongly left Hedges 0 of 20 claims Outlet record 65/100
“The articles describe different aspects of the same broader situation with OpenAI, but they report on distinct events: one about a lawsuit filed by California, and the other about OpenAI's announcement of an internal investigation.”
Toronto Star
Leans strongly left Hedges 2 of 4 claims Outlet record 64/100
“Article A reports on a lawsuit filed by California against OpenAI, while Article B discusses OpenAI's announcement of an internal assessment of its AI technology. These are different events, though they may be related to the same broader issues.”

Story

OpenAI Assessment Criticism
Technology · 1 article(s) covering the same event.
How this article was measured signals, leaning charts, publisher record, byline, topics and claims

Signals How these are calculated →

Claims extracted
21
sentences that state something as happening or true
Uncertain
10%
2 of 21 hedged ("reportedly", "may", "according to")
Leaning
Leans strongly left
of the writing, not the subject · beta estimate
Publisher signals
60/100
Mother Jones's corrections and hedging, across everything collected from it. Not a measure of accuracy.
Outlets on this story
1
Technology
Same framing elsewhere
1
articles that frame the event the way this one does
Analyzed 2026-11-17 · how these are computed

How this is being covered How these are calculated →

Article leaning vs. publisher reliability
Source leaning vs. consistency

Publisher

Mother Jones · 571 article(s) · 1 correction(s) detected
Running correction rate · 1 correction(s)

Who wrote this

Alex Nguyen
57 article(s) here · 1 carrying a prediction
Amid months of security debacles and plummeting approval of its AI technology, OpenAI announced on Monday that four independent assessors will investigate its internal models and test the effectiveness of its safety features.
President Trump, who has repeatedly defended his friend following reports of the personal scandals, told reporters Monday that he “would not mind” if Jensen Huang were reinstated as CEO, saying he thinks Huang had “learned a big lesson.”
2026-11-16 · assertive framing · Inside the Fight to Bring Jensen Huang Back to Nvidia
He also said that these disruptions could happen again until all copper wires have been replaced, which is scheduled to be completed by the end of 2028.
Rallies also took place on Wednesday in six other cities—Los Angeles; Chicago; Boston; Portland, Oregon; Boulder, Colorado; and Durham, North Carolina—organized by a nationwide alliance of parent advocacy organizations. Just over a week earlier, on Election Day, Google announced that it would turn off its AI tools for school systems unless administrators choose to turn them on.
Will Scharf, the White House counsel, wrote to remind incoming House Oversight Committee Chair Rep. Robert Garcia (D-Calif.) about what he called the Constitution’s “foundational limits.”
- Xi summit: In May of this year, Trump reportedly invited Huang to join him while he talked with President Xi about “open[ing] up” the country to US businesses.
2026-11-10 · mixed framing · Trump Is Firmly in Jensen Huang’s Corner
How will Stephen Miller, the public face of some of President Donald Trump’s cruelest policies, operate in his new role as White House chief of staff?
2026-11-09 · assertive framing · It’s Stephen Miller’s First Day as Wannabe Mob Boss
One of Election Day’s most surprising results still isn’t final: the extraordinarily close Mississippi Senate race between Republican incumbent Cindy Hyde-Smith and Democrat Scott Colom, who could become the state’s first Black senator since Reconstruction.
2026-11-04 · mixed framing · Mississippi’s Unbelievable Blue Turn
I thought it would be appropriate to have the media here.
El-Sayed, an epidemiologist and public health administrator pushing “Medicare for All” and endorsed early in the race by Sen. Bernie Sanders (I-Vt.) and Rep. Alexandria Ocasio-Cortez (D-N.Y.), ran a campaign that was a key test for Democrats—could a more progressive, less-established candidate win against a Trump-endorsed, long-serving Republican member of Congress?
2026-11-04 · mixed framing · Abdul El-Sayed Wins in Michigan
Also by Alex Nguyen
Nothing else under this byline is closely related to this article, so these are simply their most recent.
All 57 articles by Alex Nguyen →

Topics

Apollo Research Goodfire METR OpenAI Transluce

Subjects

OpenAI ORG · 13× METR ORG · 5× Apollo Research ORG · 2× Goodfire ORG · 2× Transluce ORG · 2× Mother Jones ORG · 1× Ryan Greenblatt PERSON · 1× Sam Altman PERSON · 1× The Center for Investigative Reporting ORG · 1×

Narrative

The details of Goodfire’s assessment are the most vague, with the AI research lab stating that it will have “deep access to OpenAI’s internal models and systems”—whatever “deep access” means—to use its “interpretability platform to monitor training runs, debug undesired behavior, and improve alignment.”
framing: assertive · carried by 1 article(s) · first seen 2026-11-17
Amid months of security debacles and plummeting approval of its AI technology, OpenAI announced on Monday that four independent assessors will investigate its internal models and test the effectiveness of its safety features.
2026-11-17 · Mother Jones
OpenAI’s “Independent Assessment” Doesn’t Look Independent at All · assertive framing

Claims (21 extracted, 2 hedged)

Amid months of security debacles and plummeting approval of its AI technology, OpenAI announced on Monday that four independent assessors will investigate its internal models and test the effectiveness of its safety features. asserted
assessors → plummet → features
It may sound like a step in the right direction, but there’s a glaring problem: The evaluators have limited access to the startup’s technology, and—in many cases—OpenAI gets to approve each member of the research teams and can choose which staff they speak to. uncertain
they → sound → staff
Many AI researchers have pointed out since yesterday’s announcement that the assessors don’t appear to have the “employee-like access” that CEO Sam Altman committed to in September. asserted
Altman → point → September
The four assessors—METR, Apollo Research, Transluce, and Goodfire—are research labs that evaluate AI models for safety risks. asserted
that → evaluate → risks
Two METR staff members, along with one researcher from another nonprofit AI safety research organization, investigated OpenAI in the wake of the July Hugging Face security incident. asserted
members → investigate → incident
The researchers initially had only two days to work on-site and were later invited two more times—which amounted to a total of six days. asserted
which → have → days
Following the publication of the team’s findings, one of the researchers admitted a lack of “good approaches for understanding/overseeing the activity and aims of AI” when multiple agents work together, as in the Hugging Face breach. asserted
agents → follow → breach
“Outputs from analysis [AI] agents were often missing key details, wrong, overconfident, or really hard to understand,” the researcher, Ryan Greenblatt, wrote on X in August. asserted
researcher → miss → August
The researchers’ report also acknowledged that the volume of material under review meant they “had to heavily delegate our analysis to AI agents,” which have “significantly worse judgment and reliability than human researchers.” asserted
which → acknowledge → researchers
It’s not exactly a complete study. asserted
It → ’ → ?
OpenAI said Monday it would allow METR to conduct its assessments with undefined “agreed access, resources, and safety arrangements.” asserted
METR → say → access
In a footnote, the startup acknowledged that the scope of the investigation would be “less expansive than what METR outlines in their alignment assessment proposal,” which in October attempted to outline a comprehensive method to understand the worst plausible outcomes of a developer’s AI agents. asserted
which → acknowledge → agents
Although Apollo Research, another assessor, is set to “work within OpenAI for four months,” its announcement says that OpenAI can withhold information because of “technical infeasibility, security risks, or IP protection” and will select the staff members that assessors can speak with. asserted
assessors → set → that
Similarly, Transluce can only access models, safety documentation, and researchers pre-approved by OpenAI. asserted
Transluce → access → OpenAI
The details of Goodfire’s assessment are the most vague, with the AI research lab stating that it will have “deep access to OpenAI’s internal models and systems”—whatever “deep access” means—to use its “interpretability platform to monitor training runs, debug undesired behavior, and improve alignment.” asserted
access → state → alignment
Beyond the obscure language, OpenAI and the research labs’ framing of the assessments as “collaborations” that can work toward “stronger shared standards—emphasis on “shared”—is perhaps the most telling asserted
that → work → standards
A truly independent investigation of OpenAI would not include OpenAI having final say on much of what is being investigated, or being able to dictate different terms to each assessor, with deeply different levels of time and access. asserted
what → include → time
As METR itself said in October regarding how to evaluate unreleased AI models, independent experiments would “ideally have codebase and infrastructure access comparable to what an employee studying similar questions would typically have,” including training code and the company’s simulation tools. asserted
employee → say → code
While the announcement suggests that OpenAI no longer feels it can just selectively disclose its own models’ “unexpected or concerning behavior,” it’s a far cry from the firm’s stated goal of “public transparency across the industry.” Disclosure: uncertain
it → suggest → industry
The Center for Investigative Reporting, the parent company of Mother Jones, has sued OpenAI for copyright violations. asserted
Center → sue → violations
OpenAI denies the allegations. asserted
OpenAI → deny → allegations
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