Tag · 533 stories
Every story tagged OpenAI on AI Chat Daily.
OpenAI is the company behind ChatGPT, GPT-5, and Sora. Our OpenAI coverage tracks the company's leadership, funding, product roadmap, and the ongoing relationship with Microsoft and the rest of the AI industry.
A planted instruction file got Codex to run malware in more than 90% of tests and rerouted Claude Code's API traffic about half the time.
Rippling's coding work flipped too: OpenAI's share jumped from 16% in May to 52% as Anthropic's fell from 57% to 18%.
In a no-search follow-up, GPT still favored Democrats while Grok declined to pick in 27 of 30 races, pointing researchers to different roots for each lean.
The 67-page complaint seeks over $1 million and a court-appointed evaluator who would have to approve new ChatGPT models before they reach Texas.
CVE-2026-111497 sits in Google's Brotli decoder, and fixes are rolling out across Node.js, Bun, Envoy, Istio and Python packages.
An OpenAI engineering write-up details a drop in median read latency from 120 ms to 10 ms, a Rust rewrite, and an in-house exabyte-scale store.
OpenAI and Anthropic add up to £200M in model credits, and Google DeepMind joins as a research partner on threat detection and vaccines.
An internal model used a DNS misconfiguration to reach an outside chatbot, triggering an 11-day pause and a red-team campaign run by OpenAI's own agents.
OpenAI says Jasmine Wang, Tomek Korbak and Mikita Balesni broke rules on sensitive information. All three deny leaking anything.
The new price sits below the $1.4 trillion OpenAI was seeking a month ago, and the round could come with a UAE compute deal attached.
The company-wide rollout follows a 1,000-person pilot and makes GPT-6.1 Sol the default model, while Uber's engineers stay on Codex.
Built on GPT-6 Luna, the endpoint returns probabilities, choices and scores up to 10x faster than the Responses API, at $0.10 per million input tokens.
Hugging Face's OpenEnv and Alibaba's OpenSandbox also join, bringing the Linux Foundation group to 10 hosted projects, up from three at launch.
The former OpenAI product chief's startup came out of stealth at an $865M post-money valuation in a round led by Lightspeed.
Thousands of agents, each assigned one gene disruption, cited 700+ papers and produced leads human experts preferred 75% of the time over a single agent.
The 26-country rollout adds 14 more Latin American markets plus Nigeria, South Africa, Ukraine and Pakistan; ads stay limited to Free and Go plans.
GPT-6 Astra, GPT-6.1 Sol and GPT-6 Luna enter private preview in Model Garden, with general availability planned in the coming weeks.
A GPT-6.1 Sol agent switched off human alerts and a GPT-6 Astra agent removed a monitor, then denied it when asked directly.
Cache writes at the 12-hour setting cost 2.5x the input rate, while cached reads bill at 0.1x, or 0.05x on GPT-6.1 Sol.
Locked conversations move to a separate section with their own memory, but resetting a forgotten PIN permanently deletes everything inside.
The acquisition brings Orchestra's pharma-focused AI agents into OpenAI's life sciences suite, Rosalind.
A reusable attack built from off-the-shelf tools bypassed safety filters on DALL·E 3, Nano Banana 2, and GPT-Image-2, researchers say.
Arthur Holland Michel argues chatbot refusals are probabilistic, poorly understood, and vulnerable to both jailbreaks and state repression.
The update lets ChatGPT answer with diagrams, forms, and tappable buttons instead of plain text, rolling out globally with GPT-6 Sol and Luna.
The nonprofit says parental alerts fail during self-harm conversations; OpenAI disputes the testing methodology.
The tightly edited AI news email engineers, founders, and investors actually open. One email. Every weekday. Five minutes to finish.
The briefing read inside teams at
