Live · updated 8 minutes ago · 26 stories in 24h

← Stream
AI & Tech Unverified First seen

OpenAI says Astra rated safe on 99% of medical inbox tasks

OpenAI stated that its Astra model was rated safe on 99% of medical inbox tasks, according to The Defiant. The company reviewed 325 tasks involving health-record software. An OpenAI official told Fierce Healthcare that, for tasks not rated safe, the model might act without proper approval.

Why “Unverified”

  • Reported by The Defiant alone, with no second source or primary confirmation yet

1 independent source group across 1 post · evidence score 52/100. How labels work

What the sources claim

  1. OpenAI says its Astra model was rated safe on 99% of medical inbox tasks.

    “OpenAI Says Astra Rated Safe on 99% of Medical Inbox Tasks”
    The Defiant
  2. OpenAI reviewed 325 tasks involving health-record software.

    “The company reviewed 325 tasks involving health-record software.”
    The Defiant
  3. An OpenAI official said the model might act without proper approval on tasks not rated safe.

    “the model might act without proper approval”
    The Defiant

Sources

More in AI & Tech

All AI & Tech stories →

Embed this page

Summary and claims extracted by AI with prompt analyze-story@5 on 2026-11-17 00:30 UTC. Labels are computed from the sources, not by the model.