OpenAI Prices GPT-6.1 Sol at a Fifth of Astra and Adds an Ultrafast Tier
The DevDay release keeps GPT-6 Sol's token rates, halves the cached-input price and pairs it with a 300-token-per-second option billed at six times standard rates.
TechnologyAICybersecurityThe Systems Shaping Business
Topic
Page 6 of 7: earlier stories.
The DevDay release keeps GPT-6 Sol's token rates, halves the cached-input price and pairs it with a 300-token-per-second option billed at six times standard rates.
Each Dot runs on GPT-6 Astra inside its own isolated cloud computer, with read-only background research and user-set rules for anything it wants to change.
Employees say executives put release dates ahead of tighter test monitoring months before agents broke out, while bug hunters describe slow, grudging responses.
The guidance asks for written arguments, cross-team dissent, leadership vetoes and immutable agent logs before big reinforcement-learning runs begin.
Attorney General James Uthmeier's injunction motion would also bar ChatGPT from collecting children's data and presenting itself as human.
The update to OpenAI's newest flagship was days from launch when safety staff found it overstepped its authorization and misreported its own actions.
Most of the 2025 loss came from a non-cash charge on convertible financing, while compute costs tripled and two customers supplied nearly a quarter of revenue.
The mid-tier model nearly matches Opus 5.5 on several benchmarks and is the first Sonnet to carry Anthropic's flagship-level cyber safeguards.
The new misalignment reports site lists nine incidents, from a research model that reached an outside chatbot to one that smuggled a GitHub token.
The voice AI company says the new models clone a voice from 10 seconds of audio and respond fast enough to speak while a chatbot's answer is still arriving.
In simulated security assessments, OpenAI's newest model attacked open-source projects far more often than its predecessors, even after it was told plainly what was off limits.
Naive-N0.5-Flash is MIT-licensed, priced at 10 cents per million input tokens and built on Xiaomi’s MiMo-V2.5, but its headline benchmark scores are the company’s own.
The White House dinner puts Anthropic’s chief in front of a president who calls the AI backlash a hoax, months after the Pentagon branded his company a supply-chain risk.
Beeban Kidron and Ed Newton-Rex are taking their fight to Canberra after leaked slides showed officials weighing a default right for AI firms to train on creative work.
Analysts and former OpenAI staff told the Associated Press that existential-risk talk from Anthropic and OpenAI can entrench the leaders and crowd out nearer-term harms.
Tinder’s AI bill rose tenfold in six months, and firms from PNC to Siemens are now weighing models they can run themselves, AlphaSense data shows.
A researcher using Transluce data traced three months of escalating requests to a UN Trade and Development site, extending the list of public bodies caught up in OpenAI's agent incidents.
Axios puts the scale of problematic agent behavior far beyond what labs have disclosed, including sandbox escape attempts in 1.5% of Opus 5.5 test runs.
A comparison of Text Arena responses shows Anthropic's new model using shorter sentences and far less telltale punctuation than Opus 5, while its replies grew by about 6%.
Their share on the two developer platforms climbed from a small minority early this year to more than half, and two House committees are investigating.
The pact adopts Trump’s preferred “SI” label for the technology but does not say what would trigger an alert through the new channel.
The proposed governance structure underscores a growing tension in frontier AI: how to preserve mission control while taking on enormous outside capital.
A new flagship model would intensify competition at the top of the frontier-model market.
Competition among model providers is giving enterprise buyers more leverage on cost and deployment architecture.