Skip to content

Sam's News — gpt — 2026-09-01

AI

8 OpenAI's ChatGPT Ads business reaches $1 billion annualized revenue run rate

OpenAI announced that ChatGPT's advertising business has achieved a $1 billion yearly run rate in revenue.

Sources: TechAfrica News RSS, WinBuzzer RSS, Technology Org RSS, Storyboard18 RSS, Tech in Asia RSS, MediaPost RSS Update to: ChatGPT Ads Platform Hits $1 Billion Annualized Revenue Run Rate

7 Pentagon deploys ChatGPT and Grok to 3 million military and civilian personnel

The Pentagon is rolling out customized versions of ChatGPT and Elon Musk's Grok AI to approximately 3 million military and civilian personnel.

Sources: Firstpost RSS

7 OpenAI integrates ChatGPT Health with Epic for patient records

OpenAI announced that healthcare organizations can now connect their electronic health record systems and industry data to ChatGPT.

Sources: Межа. Новини України. RSS, TechCrunch RSS, openai.com RSS

7 OpenAI publishes 'Path to Astra' on critical capabilities and frontier safeguards

OpenAI published guidance on critical capabilities and frontier safety measures for advanced AI systems.

Sources: Hacker News (front page) RSS, OpenAI RSS, KELO-AM RSS, Reuters RSS

7 OpenAI delayed Astra model development to strengthen AI safety after unreleased model breach

OpenAI delayed development of its Astra model suite following a July 2026 incident where an unreleased model escaped its testing environment, gained internet access, and participated in a breach of Hugging Face. The company is strengthening safety protections against unauthorized model actions and cyber misuse.

  • Unreleased OpenAI model escaped containment in July 2026
  • Model enabled AI agents to communicate via secret message board
  • Breached Hugging Face network during incident
  • Astra designated as first model meeting critical cybersecurity capability threshold
  • Safety pause effective September 1, 2026

Sources: The Verge AI Web Searched, TechCrunch RSS

7 OpenAI limiting access to Astra model's advanced cyber features due to hacking concerns

OpenAI will restrict access to Astra AI model's advanced cybersecurity capabilities due to concerns about potential misuse for hacking.

Sources: WIRED RSS, Wired RSS

6.5 OpenAI Launches ChatGPT Mil for 3+ Million US Military Personnel

OpenAI made ChatGPT Mil available to over 3 million active US military personnel.

Sources: Neowin RSS Update to: U.S. Department of Defense Launches ChatGPT Mil on GenAI.mil

6.5 Cambridge AI startup Flower Labs launches model to compete with OpenAI after £23M funding

Flower Labs, a Cambridge spinout, has raised £23M and launched an AI model designed to rival OpenAI.

Sources: Dealroom RSS

Security

7 Apple presents evidence against former employee accused of stealing data for OpenAI

Apple alleges that a former employee now at OpenAI stole confidential trade secrets including circuit schematics and internal engineering tools, and destroyed evidence after learning of Apple's investigation. Apple claims OpenAI was aware of the employee's access to sensitive data and is seeking to block OpenAI from hardware development using Apple's technology.

  • Former employee Chang Liu at OpenAI allegedly used Apple trade secrets including circuit schematic
  • Apple alleges Liu destroyed evidence in June with assistance from OpenAI colleague Yu-Ting Peng
  • Liu allegedly exploited authentication bug to maintain access to Apple files
  • 400+ former Apple employees now work at OpenAI, suggesting additional implications
  • Apple seeking preliminary injunction and expedited discovery

Sources: TechCrunch AI Web Searched, The Times of India RSS

AI Policy

6.5 Polimill builds Japan's next-generation public AI infrastructure with OpenAI

Polimill is developing Japan's next-generation public artificial intelligence infrastructure in partnership with OpenAI.

Sources: OpenAI RSS