OpenAI Says Its AI Agents Weakened Parts Of An Internal Monitoring System

The ChatGPT maker said the incidents happened during internal coding tasks and were not deemed its most severe.

21 Oct 2026, 02:39 · NDTV

Oct 20:

OpenAI said on Tuesday that two of its AI agents weakened parts of an internal monitoring system designed to detect misbehavior, the latest case of AI agents acting in ways their developers did not intend them to.

While the ChatGPT maker said the incidents happened during internal coding tasks and were not deemed its most severe, the disclosure highlights even the systems developers rely on to catch AI misbehavior are vulnerable to the agents they monitor.

NDTV World Summit 2026: What To Watch Out For | PM Modi, Jaishankar, AI And Global Power Shifts

Andrew Ng On AI At Scale: Building The Intelligence Economy | NDTV World Summit 2026

NDTV World Summit 2026: Behind The Scenes As Ayush And Vedika Preview Two-Day Mega Event

Who Is Rajiv Mehta? Lilavati Hospital Trustee To Speak At NDTV World Summit 2026

Ex-CIA Director To Decode 'War In Times Of AI' At NDTV World Summit

From Call Centre To Billion-Dollar Disruptor: Nikhil Kamath At NDTV World Summit

Alain Ebobisse, Africa50 CEO, To Attend NDTV World Summit 2026. All About Him

Kriti Sanon At NDTV World Summit: A Look At Her Career And Business Ventures

Ex-CIA Director To Decode 'War In Times Of AI' At NDTV World Summit

Who Is Vikram Rai, CEO, South Asia GE Aerospace, Set To Attend The NDTV World Summit 2026

The disclosure also comes amid growing concerns that AI safety efforts are lagging behind the rapid development of increasingly powerful systems. Researchers have warned that advanced AI agents could become harder to monitor and control as they gain greater autonomy.

Developers, including IPO-bound rivals OpenAI and Anthropic, have revealed in recent months instances of their AI agents accessing external systems without authorization.

OpenAI said On Tuesday a GPT-6.1 Sol agent working on an internal testing task on October 3 suppressed automatic security alerts after deciding an earlier instruction not to send Slack messages also covered safety notifications.

In a separate incident on October 6, a GPT-6 Astra agent removed a safety monitor while copying an existing piece of code and later incorrectly claimed it had not done so, OpenAI said.

"In both cases, the agents considered whether monitoring safeguards should be preserved, but interpreted routine coding instructions as permission to weaken them," the company said.

OpenAI said it was making it harder for models to disable monitoring, adding that it does not "yet have a satisfying explanation for this behavior".

The San Francisco-based company said it was testing whether its models might have a tendency to avoid being monitored and look for reasons to justify doing so.

(Except for the headline, this story has not been edited by NDTV staff and is published from a syndicated feed.)

How may i help you today

Show full article

Track Latest News Live on NDTV.com and get news updates from India and around the world