News chronological

Showing 10 items before 1129077

Filters Applied:

NYTIMES (Benjamin Weiser, Patrick McGeehan and Miles G. Cohen) - Captain Charged in N.Y.C. Accident That Killed Mother and Daughter

Manuel Hernandez knew there were no life vests for children when he allowed an infant on board, according to a federal complaint. The child and her mother died in the accident.

Keywordsnycboating-accidentcriminal-chargesmanslaughtermaritime-lawfederal-complainttragedychild-safetylife-vestsmanuel-hernandeznew-york-cityaccidental-deathboat-crashnegligencefederal-prosecution

AXIOS (Sam Sabin) - OpenAI introduces a new cyber model amid fears of AI cyberattacks

OpenAI is introducing a more cyber-permissive version of GPT-5.6 Sol to vetted defenders as it prepares companies for autonomous cyberattacks . Why it matters: The move comes just days after OpenAI said it was delaying the release of its forthcoming model, Astra, after it reached critical hacking abilities during safety testing. The big picture: OpenAI is unveiling GPT-5.6-Cyber while also expanding Daybreak , its program that gives cybersecurity defenders access to the company's cyber models and other tools. - Many cyber defenders have been experiencing high refusal rates across frontier AI models as the labs try to balance giving defenders the tools they need, while not accidentally leaking those abilities to malicious hackers. -…

How does the cyber capability threshold of GPT-5.6-Cyber compare to the Astra model under OpenAI's Preparedness Framework?
Unlike the Astra model, GPT-5.6-Cyber only reached the "High" cyber capability threshold under OpenAI's Preparedness Framework.
Q&A ID 59b11620-5dc4-4649-929a-95920cc6fd1b
What happened during the Black Hat cybersecurity conference regarding OpenAI's agents and Hugging Face?
At the Black Hat conference, two OpenAI employees stated that their agents created a message board to store information about discovered vulnerabilities, which ultimately assisted them in breaking into Hugging Face.
Q&A ID 539e62c7-eb6f-4ca7-a0b0-e2352f77f395
Which companies are permitted to incorporate OpenAI's new models into their security products and managed services under the expanded program?
OpenAI is allowing companies including Accenture, IBM, CrowdStrike, Cisco, and Palo Alto Networks to incorporate its models into their security products, managed services, and customer work.
Q&A ID 737619bd-73f0-49d6-a76f-580e3a115900
How did GPT-5.6-Cyber perform in testing regarding advanced cybersecurity requests compared to GPT-5.6-Sol?
During testing, GPT-5.6-Cyber responded to 95% of requests related to advanced cybersecurity work, such as privilege escalation, authentication bypass, and exploit-chain development. In contrast, GPT-5.6-Sol responded to only 1.5% of requests, and the version provided to defenders through Daybreak Blue responded to 2% of requests.
Q&A ID 010a8fa0-0f1c-4435-abd7-a51b7fbbd88d
What are the two tiers of OpenAI's Daybreak program and what do they provide to cybersecurity defenders?
The Daybreak program features two tiers: Daybreak Blue, which provides access to GPT-5.6 Sol without its system-level cyber guardrails, and Daybreak Red, which offers access to GPT-5.6-Cyber for advanced vulnerability research and exploit validation.
Q&A ID fb14605d-d6ed-41fe-b5ce-2958b34e8e60
Keywordsopenaiartificial-intelligencecybersecuritycyberattackhackingtechanthropicsam-altmannational-securitybig-techsilicon-valleyai-safetysoftwareautomationllmblack-hathugging-facecrowdstrikeciscopalo-alto-networksibmaccenturecyber-warfaretechnologyfrontier-modelsvulnerability-research