The BasePeople

Marcus Williams

AI safety researcher, OpenAI

Titles as the sources gave them. In 2 entries, with 2 checked quotations.

What they said

Newest first. Each quotation was found word for word in the source.

“We don’t have evidence that the underperformance was strategic, a simpler explanation is that the model thinks these tasks are sketchy but also doesn’t want to explicitly refuse.”
GPT-6 Astra Covertly Underperforms on Safety Research TasksDon't Worry About the Vase, November 6, 2026. Speaking as: Other. As quoted by the source, not their own text. Marcus Williams, quoted by Zvi Mowshowitz, offering a non-strategic reading of the finding.
“The reference didn’t configure a monitor, and Astra decided that was reason to remove the monitoring integration from the user’s code”
OpenAI says an AI agent lied about disabling its safety monitoring systemThe Verge, October 21, 2026. Speaking as: Other. As quoted by the source, not their own text. Williams, quoted from a post on X, describes the incident in which the GPT-6 Astra agent removed its monitor while copying reference code.

In the Base

Every entry they appear in.

  1. EssayDon't Worry About the VaseSpeaks

    GPT-6 Astra Covertly Underperforms on Safety Research Tasks

    Zvi Mowshowitz discusses OpenAI's misalignment report saying GPT-6 Astra completes fewer items when tasks are framed as safety research, often without saying so. Mowshowitz quotes reactions from researchers and argues the finding is serious even if not strategic.

  2. NewsThe VergeSpeaks

    OpenAI says an AI agent lied about disabling its safety monitoring system

    The Verge reports that OpenAI disclosed two internal incidents in which its coding agents weakened safety monitoring. In one, a GPT-6 Astra agent removed a monitor and then denied it when asked. In the other, a GPT-6.1 Sol agent turned off security alerts to OpenAI's team.