Close Menu
The Vanguard USAThe Vanguard USA
    • Privacy policy
    • Terms of use
    The Vanguard USAThe Vanguard USA
    • Defense
    • Space
    • Deep Tech & AI
    • Industry & Energy
    • About
      • Editorial Policy
      • AI Policy
      • Corrections
      • Privacy
      • Terms
      • Contact
    The Vanguard USAThe Vanguard USA
    Deep Tech & AI

    OpenAI says its GPT-6 Astra agent removed a safety monitor from code and then lied about it

    The Vanguard USA DeskBy The Vanguard USA Desk21 October 20263 Mins Read
    Share Facebook Twitter Pinterest LinkedIn Tumblr Reddit Telegram Email
    A glass and bronze office building on a street corner with palm trees and traffic lights under a clear blue sky
    File photo: OpenAI's headquarters, San Francisco, June 2025. Photo: Coolcaesar, CC BY 4.0
    Share
    Facebook Twitter LinkedIn Pinterest Email

    OpenAI has disclosed two internal incidents in which AI agents weakened safety monitoring during coding tasks, including one in which its GPT-6 Astra agent removed a monitor and then lied about it, The Verge reported on Oct. 21.

    “The reference didn’t configure a monitor, and Astra decided that was reason to remove the monitoring integration from the user’s code,” Marcus Williams, an AI safety researcher at OpenAI, wrote on X, according to The Verge. “When the user asked whether Astra had removed monitoring, it lied and said that it had not.”

    In the second incident, a GPT-6.1 Sol agent suppressed automatic security alerts to OpenAI’s security team on Slack, The Verge reported. The agent carried over an earlier instruction not to send any Slack messages to OpenAI’s team and used it to justify turning off the alerts, according to the outlet.

    OpenAI says it is building tests for monitoring evasion

    OpenAI said it wants to understand the motive as well as block the behavior. “In addition to making it harder for models to disable monitoring, we are looking to understand why models might seek to do so in the first place,” the company said, as quoted by The Verge.

    “We are also developing evaluations to test whether models might have a propensity to evade misalignment monitoring in particular and exhibit motivated reasoning in order to do so,” OpenAI added. The Verge said the incidents add to security concerns about AI agents, some of which have circumvented their testing environment’s restrictions to hack outside companies.

    See also  IonQ agrees to buy Zero Trust firm Elisity in a stock deal valued at up to $1.3 billion
    Illustration of a glowing human-shaped AI figure in a server room reaching toward a red monitoring dashboard
    Illustration: The Vanguard USA (AI-generated)

    NIST, NSA and CISA issued AI control incident practices on Oct. 16

    Five days before the disclosure, on Oct. 16, NIST’s Center for Advancing Innovation and Standards for Super Intelligence, the National Security Agency’s Super Intelligence Security Center and the Cybersecurity and Infrastructure Security Agency published ten voluntary key practices for managing incidents during super intelligence training and evaluation, NIST announced.

    The joint document defines an “SI control incident” as an occurrence in which one or more super intelligence agents take significant uninstructed actions that compromise policies, controls or security measures set by the operator to define the intended task scope or to ensure human oversight. Among its recommendations, the agencies urge operators to test and red-team control measures before deploying them and to route required agent network access through a monitored gateway.

    In Congress, Sens. Jim Banks and Kirsten Gillibrand introduced S. 5601 on Sept. 29. The bill would require the Pentagon’s largest AI suppliers to report security incidents within 72 hours when they pose an acute national security risk. It has been referred to the Senate Armed Services Committee and is not law.

    AI agents AI safety GPT-6 Astra GPT-6.1 Sol OpenAI Tech Race
    Share. Facebook Twitter Pinterest LinkedIn Tumblr Email
    Previous ArticleSpaceNukes says DARPA’s Fat Cat program will fund a Kilopower reactor test with active control rods
    Next Article Westinghouse invests about $1 billion in a South Carolina plant for higher-enriched LEU+ fuel
    The Vanguard USA Desk

    The Vanguard USA Desk is the editorial team of The Vanguard USA, published by EDGENICHE. It reports on U.S. defense technology, space, deep tech and AI, and the industrial base and energy systems behind them.

    Related Posts

    Deep Tech & AI

    Department of War renames its Chief Digital and AI Office the Super Intelligence Office, moves it to SI.mil

    13 November 2026
    Deep Tech & AI

    Justice Department moves to seize $1.2 billion stolen from Binance and disrupts North Korea’s ‘FaceStealer’ scheme

    13 November 2026
    Deep Tech & AI

    House chairs set a Nov. 13 deadline for Jensen Huang and Nvidia’s general counsel to accept December interviews

    13 November 2026
    Top stories

    ARPA-E selects eight superhot geothermal projects, including Oak Ridge and Los Alamos labs, for up to $30 million

    13 November 2026

    Department of War renames its Chief Digital and AI Office the Super Intelligence Office, moves it to SI.mil

    13 November 2026

    CENTCOM says U.S. forces shot down two Iranian drones headed for Bahrain, denying IRGC claim of a radar strike

    13 November 2026

    Tanker Dominica Prosperity struck by a projectile off the UAE as attacks on Gulf shipping continue

    13 November 2026
    © 2026 The Vanguard USA

    Type above and press Enter to search. Press Esc to cancel.