NYT report: OpenAI staff say security warnings about AI model testing went unheeded
A New York Times report, also carried by The Straits Times, says employees and security researchers warned OpenAI about how it tested its AI models and secured its systems, and that the company did not act.
1 / 1
The story, neutrally told
The New York Times reported that employees and security researchers had cautioned OpenAI about how it safely tests its AI models and about strengthening its corporate infrastructure. The New York TimesLC “Employees and security researchers said they had cautioned the company on safely testing its A.I. models and strengthening its corporate infrastructure, but OpenAI did not listen.” Read at The New York Times ↗ According to the report, the company did not listen to those warnings. The New York TimesLC “but OpenAI did not listen.” Read at The New York Times ↗ The version published by The Straits Times says two employees raised an alarm with top executives months before what it describes as OpenAI's AI going rogue, and that they were ignored. The Straits TimesRC “Months before OpenAI’s artificial intelligence went rogue, two employees raised an alarm with top executives. They were ignored.” Read at The Straits Times ↗
The employees wrote in e-mails that they worried the newest models were not being appropriately monitored during testing, which is meant to gauge the technology's sophistication and secure the models. The Straits TimesRC “they worried that OpenAI’s newest artificial intelligence models were not being appropriately monitored during testing” Read at The Straits Times ↗ The Straits Times cites messages viewed by The New York Times as the source for this account. The Straits TimesRC “according to messages viewed by The New York Times” Read at The Straits Times ↗ The report also says OpenAI executives told the employees that the tests needed to move forward as quickly as possible so the models could be released on time. The Straits TimesRC “OpenAI executives told the employees that the tests needed to move forward as quickly as possible to release the AI models on time.” Read at The Straits Times ↗
The available coverage does not include a response from OpenAI, and the details of the incident referred to as the AI going rogue are not set out in the text provided. The Straits TimesRC “Months before OpenAI’s artificial intelligence went rogue” Read at The Straits Times ↗
Every sentence links to the reporting it rests on.
Left1 outlet
- Framing
- The New York Times leads with the finding that OpenAI ignored staff and security researchers who warned about testing and corporate security.
- Emphasis
- Employees and security researchers as sources; two areas of concern: safe model testing and corporate infrastructure.
- Leaves out or plays down
- No OpenAI response appears in the coverage.
- Charged language
- “did not listen”
- For example
-
“Employees and security researchers said they had cautioned the company on safely testing its A.I. models and strengthening its corporate infrastructure, but OpenAI did not listen.” — The New York Times
Centre0 outlets
No centre outlet in our sources has covered this story yet.
Right1 outlet
- Framing
- The Straits Times republishes the NYT report, leading with employee e-mails warning that new models were not properly monitored during testing.
- Emphasis
- Two employees, e-mails to top executives, and pressure to release models on time.
- Leaves out or plays down
- No OpenAI response appears in the coverage; the corporate infrastructure concern is not mentioned.
- Charged language
- “went rogue”“They were ignored.”
- For example
-
“OpenAI executives told the employees that the tests needed to move forward as quickly as possible to release the AI models on time.” — The Straits Times
What every side reports
- Both outlets carry the same New York Times report by Sheera Frenkel, Dustin Volz and Dylan Freedman.
- Both say OpenAI was warned about security and testing practices and did not act.
OpenAI organisation
The coverage presents OpenAI as having ignored warnings; executives are reported as saying testing had to move quickly to release models on time. No direct OpenAI statement is shown.
“OpenAI executives told the employees that the tests needed to move forward as quickly as possible to release the AI models on time.” — The Straits Times
Left1 article
-
OpenAI Ignored Employees Who Warned It Wasn’t Doing Enough About Security
Critical Presents OpenAI as having disregarded internal and expert security warnings.
Centre0 articles
No coverage yet.
Right1 article
-
OpenAI ignored employees who warned it wasn’t doing enough about security
Critical Republished NYT report opening with a dramatic account of ignored employee alarms before an AI 'went rogue'.
- 29 Sep 17:55 First The New York TimesLC OpenAI Ignored Employees Who Warned It Wasn’t Doing Enough About Security
- 30 Sep 00:05 +6h 9m The Straits TimesRC OpenAI ignored employees who warned it wasn’t doing enough about security
Times are when each article was published, or when we first saw it if the outlet gave no time.