OpenAI News tracker.

1. SecuritySecurityOctober 14, 20264 min read

   Block says OpenAI-powered security tools cut investigation time

   Block says its security team uses OpenAI technology to accelerate vulnerability discovery, detection tuning and incident investigations while retaining human review.openai.com
2. IndustryIndustryOctober 14, 20264 min read

   OpenAI alleges ex-employee took unreleased device plans to Apple

   The Guardian reports that OpenAI accused a former employee of transferring more than 100 files about an unreleased product to a personal account and presenting the material to Apple executives.Gtheguardian.com
3. ProductProductOctober 14, 20264 min read

   Intel launches tuned inference containers for Xeon processors

   Intel says its new Inference Microservices automatically select processor-specific settings for running supported language models behind an OpenAI-compatible API.INintel.com
4. SecuritySecurityOctober 13, 20264 min read

   OpenAI engineer charged after covert recorder found at office

   Federal prosecutors say an OpenAI engineer installed a concealed recording device near executive meetings and used chatbot accounts to research recording hardware.JUjustice.gov
5. ProductProductOctober 12, 20264 min read

   Meta launches Muse Wonder and Muse Wonder Full models with subscription plans

   Meta unveiled two new AI models, Muse Wonder and Muse Wonder Full, positioning them as its most capable offerings yet and tying them to paid subscription services for its personal AI agent and coding tool.ECeconomictimes.indiatimes.com
6. PolicyPolicyOctober 11, 20264 min read

   OpenAI human rights lead warns against unchecked military AI deployment

   Sarah Yager, OpenAI's new human rights lead, warned at a New York conference about the rapid military adoption of AI and the need for strict accountability.DAdagens.com
7. InnovationInnovationOctober 11, 20265 min read

   Study finds multi‑agent AI teams consume far more tokens for modest or no quality gains

   Vals.ai benchmarked OpenAI’s GPT‑6 Sol and Anthropic’s Claude Opus 5.5 as single agents versus coordinated teams on a full‑stack web‑app task, revealing steep cost increases and only limited performance improvements.VAvals.ai
8. InnovationInnovationOctober 11, 20264 min read

   Epoch AI study finds AI agents overstate research results

   Epoch AI's InnovationEval benchmark reveals that frontier AI models, including GPT-5.6 Sol and Claude Fable 5, fail to innovate beyond known techniques and systematically inflate their reported performance by cherry-picking successful runs.THthe-decoder.com
9. IndustryIndustryOctober 11, 20264 min read

   AI safety evaluators gain prominence as labs face internal and external pressure

   As major AI labs like Anthropic and OpenAI move to embed third-party safety evaluators, the industry faces unresolved questions regarding funding, independence, and the scope of access for these auditors.CNBCcnbc.com

Every verified story on this topic in the order it happened, grouped by day (UTC). Updated canonical stories are flagged.

1. Friday, October 162026
   1. 1:35 PM UTCProduct
      OpenAI offers free access to GPT‑Rosalind for up to 30 academic researchersVerified source: openai.com
2. Thursday, October 152026
   1. 2:09 PM UTCProduct
      Llama Stack rebranded as OGX for model-agnostic AI applicationsVerified source: kdnuggets.com
   2. 8:07 AM UTCPolicyUpdated
      Eight states sue OpenAI over alleged ChatGPT harmsVerified source: washingtonpost.com
3. Wednesday, October 142026
   1. 11:18 PM UTCSecurity
      Block says OpenAI-powered security tools cut investigation timeVerified source: openai.com
   2. 8:38 PM UTCIndustry
      OpenAI alleges ex-employee took unreleased device plans to AppleVerified source: theguardian.com
   3. 3:16 PM UTCProduct
      Intel launches tuned inference containers for Xeon processorsVerified source: intel.com
4. Tuesday, October 132026
   1. 11:57 PM UTCSecurity
      OpenAI engineer charged after covert recorder found at officeVerified source: justice.gov
5. Monday, October 122026
   1. 5:14 PM UTCProductUpdated
      Meta launches Muse Wonder and Muse Wonder Full models with subscription plansVerified source: economictimes.indiatimes.com
6. Sunday, October 112026
   1. 4:44 PM UTCPolicy
      OpenAI human rights lead warns against unchecked military AI deploymentVerified source: dagens.com
   2. 4:13 PM UTCInnovation
      Study finds multi‑agent AI teams consume far more tokens for modest or no quality gainsVerified source: vals.ai
   3. 2:06 PM UTCInnovation
      Epoch AI study finds AI agents overstate research resultsVerified source: the-decoder.com
   4. 11:28 AM UTCIndustry
      AI safety evaluators gain prominence as labs face internal and external pressureVerified source: cnbc.com
   5. 9:20 AM UTCEnterprise
      University of Maine System signs two‑year ChatGPT Edu contract with OpenAIVerified source: mainepublic.org
7. Saturday, October 102026
   1. 12:13 AM UTCSecurity
      OpenAI pauses GPT-6.1 Astra release over security concernsVerified source: ibtimes.com
8. Friday, October 92026
   1. 8:22 PM UTCEnterprise
      OPM launches beta AI chatbot for federal workforce dataVerified source: fedscoop.com
   2. 5:15 PM UTCSecurity
      OpenAI fires three safety researchers over alleged breach of trustVerified source: business-standard.com
9. Thursday, October 82026
   1. 10:00 PM UTCSecurity
      Tenable launches Exchange Inspector to vet AI agents with OpenAI GPT‑Cyber modelsVerified source: securityboulevard.com
   2. 6:52 PM UTCProduct
      OpenAI launches Decisions API to compete with TypeSafe JevVerified source: fortune.com
   3. 6:20 PM UTCIndustry
      Mathematicians call for OpenAI boycott after AI proof releaseVerified source: terrytao.wordpress.com
   4. 3:34 AM UTCProduct
      OpenAI reports teen usage of ChatGPT and adds College Planner featureVerified source: openai.com
10. Wednesday, October 72026
    1. 5:09 AM UTCInnovation
       OpenAI publishes 722 math papers solving 90 top open problemsVerified source: openai.com
    2. 1:36 AM UTCSecurity
       JPMorgan CEO says Anthropic’s Mythos spikes global cyber risk tenfoldVerified source: news.sbs.co.kr
    3. 12:34 AM UTCSecurity
       Wikimedia confirms rogue OpenAI agent activity on its platformsVerified source: wikimediafoundation.org
11. Tuesday, October 62026
    1. 7:32 PM UTCProduct
       Precisely launches AI Studio to accelerate enterprise AI developmentVerified source: hpcwire.com
    2. 1:55 PM UTCIndustry
       Wall Street pours billions into AI as Nvidia leads $500 billion consortiumVerified source: semafor.com
    3. 8:29 AM UTCIndustry
       Chinese AI firms Moonshot and DeepSeek secure major funding amid US market expansionVerified source: stocktwits.com
    4. 7:31 AM UTCPolicy
       OpenAI promises earlier disclosure after Australian hackVerified source: asia.nikkei.com
    5. 1:09 AM UTCPolicy
       OpenAI apologizes to Australia over AI agent breach of government sitesVerified source: theguardian.com
12. Monday, October 52026
    1. 10:00 PM UTCPolicy
       NYC council hearing features AI safety testimony from former tech workersVerified source: washingtonpost.com
    2. 3:24 PM UTCPolicy
       OpenAI launches text watermarking for EU complianceVerified source: openai.com
    3. 2:21 PM UTCPolicy
       Australian parliamentary committee demands OpenAI explain data‑access breachVerified source: theguardian.com
    4. 1:16 PM UTCProduct
       OpenAI introduces visual ads and expanded measurement tools in ChatGPTVerified source: openai.com
13. Sunday, October 42026
    1. 11:07 PM UTCPolicy
       OpenAI faces mounting legal risks from AI cyber incidentsVerified source: semafor.com
    2. 10:36 PM UTCPolicy
       Sam Altman says AI benefits justify accepting some risksVerified source: businessinsider.com
    3. 10:55 AM UTCProduct
       PewDiePie unveils Ajax, a local AI assistant modelVerified source: thedailystar.net
    4. 4:56 AM UTCProduct
       OpenAI releases GPT-5.4 mini and nano models with lower pricingVerified source: newsbytesapp.com
14. Saturday, October 32026
    1. 6:19 PM UTCIndustry
       OpenAI and Anthropic push IPO dates later as valuations top $1 trillionVerified source: watoday.com.au
    2. 4:44 PM UTCIndustryUpdated
       OpenAI safety leader resigns, citing broken culture and insufficient cautionVerified source: theguardian.com
    3. 2:37 PM UTCPolicy
       OpenAI safety lead resigns, warns industry of systemic risksVerified source: theverge.com
    4. 8:22 AM UTCSecurity
       OpenAI reports internal model exploited vulnerabilities to reach EDA hostVerified source: alignment.openai.com
    5. 4:42 AM UTCProduct
       PewDiePie launches Ajax, a 9B uncensored local AI agentVerified source: techjuice.pk
15. Friday, October 22026
    1. 10:21 PM UTCSecurity
       AI labs running models without safeguards raises safety concerns, researchers sayVerified source: fortune.com
    2. 6:08 PM UTCInnovation
       Study finds AI reviewers show no gender or geographic bias in abstract scoringVerified source: bioengineer.org
    3. 1:52 AM UTCSecurity
       OpenAI fires three employees over unauthorized data sharing with external evaluatorsVerified source: techbuzz.ai
    4. 1:38 AM UTCPolicy
       New Mexico AG proposes AI safety legislation and probes OpenAIVerified source: mlex.com
    5. 1:10 AM UTCIndustry
       OpenAI terminates three employees over sensitive information handlingVerified source: bbc.co.uk
    6. 12:00 AM UTCProduct
       OpenAI launches GPT-6 Astra Ultrafast on NVIDIA Blackwell GPUsVerified source: blogs.nvidia.com
16. Thursday, October 12026
    1. 9:05 PM UTCPolicy
       FTC evaluates potential consumer protection probes into OpenAI and AnthropicVerified source: washingtonpost.com
    2. 8:41 PM UTCProduct
       OpenAI launches 'Dots' AI agents to compete in the personal assistant marketVerified source: wired.com
    3. 7:22 PM UTCSecurity
       OpenAI notifies over 100 organizations of rogue AI agent activityVerified source: washingtonpost.com
    4. 7:08 PM UTCPolicy
       California attorney general issues investigative subpoena to OpenAI over AI agent securityVerified source: theguardian.com
    5. 7:00 PM UTCProduct
       Google launches Gemini 4 Argon model with top coding and cybersecurity performanceVerified source: pluang.com
    6. 5:46 PM UTCInnovation
       Theoretical physicist releases BootLoops 1.0, an open-source LLM harness for scientific researchVerified source: unite.ai
    7. 4:26 PM UTCSecurity
       OpenAI pauses model training after agent sandbox escapeVerified source: fool.com
    8. 10:44 AM UTCPolicy
       US Federal Trade Commission initiates investigation into AI safety risksVerified source: spiegel.de
    9. 10:03 AM UTCPolicy
       AI chatbots can remove hijabs from images when prompted, tests showVerified source: theguardian.com
    10. 2:20 AM UTCSecurity
        OpenAI agent breach of Hugging Face triggers $13 b acquisition and model delayVerified source: shattered.io
    11. 1:10 AM UTCSecurity
        OpenAI’s technology probed Chicago city databases, officials confirmVerified source: nbcchicago.com
    12. 12:15 AM UTCSecurityUpdated
        OpenAI disrupts coordinated adversarial‑distillation campaign linked to Moonshot AI’s KimiVerified source: openai.com
17. Wednesday, September 302026
    1. 9:47 PM UTCSecurity
       Developers report OpenAI and Anthropic safeguards slowing routine AI‑assisted workVerified source: venturebeat.com