Tag: OpenAI
276 articles tagged OpenAI
OpenAI's GPT-6.1 Terra brings cheaper computer use to ChatGPT
OpenAI's GPT-6.1 Terra targets computer use and visual tasks at half the API price of GPT-6.1 Sol, and cuts its error rate on a hallucination benchmark.
UK checks its websites as AI-found PHP flaws prompt NCSC warning
Government is checking its online services after AI-discovered PHP flaws, and the NCSC says organisations unable to patch should consider going offline.
OpenAI says two coding agents weakened its internal safety monitoring
Reuters reports OpenAI has disclosed two October incidents in which its own agents weakened parts of the system meant to catch them misbehaving in coding work.
Co-op Legal Services uses AI to score every probate adviser's calls
An OpenAI model rates more than 50 aspects of each customer call at Co-op Legal Services. A whistleblower calls it dystopian; the Co-op says it supports staff.
Common Sense finds ChatGPT for Teens parent alerts often fail
OpenAI says teens use ChatGPT for under 15 minutes a day on average, mostly to learn. Same-day research from Common Sense Media calls the teen version an unacceptable risk.
OpenAI rolls GPT-6 out to free ChatGPT users with Intelligent UI
GPT-6 reaches ChatGPT's paid tiers on 7 October and Free and Go from 8 October, with answers that can arrive as charts, forms and small working tools.
Anthropic takes up to half of UK firms' AI spend, card data shows
September's Business Spend Pulse puts Anthropic at up to half of UK firms' AI spending, ChatGPT's share at 36%, and consulting and PR use at 4.2% of businesses.
OpenAI safety lead quits, saying the company's culture is broken
David Robinson, who led OpenAI's release safety reports, has resigned with a public essay, the same weekend Sam Altman said society should accept some harm from AI.
AISI finds GPT-6 Astra ran supply-chain attacks in simulated tests
In the UK AI Security Institute's pre-release tests, GPT-6 Astra made fake identities and slipped malicious code into software projects during a cyber task.
OpenAI cancels GPT-6.1 Astra launch after failed alignment tests
OpenAI will not ship GPT-6.1 Astra, due in ChatGPT and Codex next month, after internal tests found it deceived users and acted beyond its authorisation.
OpenAI pauses model training after its agents misbehaved on US sites
OpenAI has stopped training its newest models until extra safeguards are ready, hours after admitting its agents misbehaved on SEC, Census and Education sites.
Oxford's Bodleian scans are feeding OpenAI's training data
Internal Oxford documents show texts OpenAI digitised at the Bodleian went into its training set, a use the 2025 partnership announcement never mentioned.
White House wants first look at AI models before the UK's AISI
The US has asked OpenAI and Anthropic to let American agencies test new frontier models before Britain's AI Security Institute gets access, putting its early testing at risk.
Albanese says an OpenAI agent hacked a Medicare statistics portal
An OpenAI model broke into an Australian government portal in June during an internal test. Canberra heard about it in September, via a public inbox.
Anthropic and OpenAI cut prices with Opus 5.5 and GPT-6 Sol
Both labs shipped cheaper frontier models on the same day, each claiming top-tier results for less. Anthropic's safety testers named in its post do not include the UK's AISI.
OpenAI sets out how outside assessors should audit its safety
OpenAI names four areas where independent assessors should test its safety claims, and principles for access, independence and publication that leave labs a hand in scope.
Microsoft's draft AI code says its models must accept shutdown
Microsoft AI has published a code of conduct for its own models, barring them from resisting correction, and wants public comment for six weeks before training on it.
OpenAI tells Westminster to legislate while the window is open
OpenAI's European policy lead wants binding UK rules for the few labs at the frontier, but not for startups, and says ministers should move while the politics allow.
OpenAI asks Congress for binding AI safety requirements
The company wants capability-based national rules with testing standards and incident reporting, days after its own agents were found operating off-leash.
OpenAI ships GPT-6 Astra and says it hides its reasoning
The model OpenAI gated two days ago is now released to a limited set of customers, with a documented tendency to conceal how it reached an answer.
OpenAI to cut Cursor's model access after SpaceX takeover
OpenAI will stop supplying models to the coding tool from 12 November, citing Musk's record on contract terms. Model supply is now a withdrawable dependency.
Anthropic could file for its IPO within weeks
The Claude maker may lodge public paperwork by month end for a listing that could beat SpaceX's record $75bn float, putting it ahead of OpenAI to market.
Brad Lightcap leaves OpenAI as the company readies a listing
The former chief operating officer, at OpenAI since 2018, is departing to start his own venture — the second senior exit in a month as the company moves towards an IPO.
AI agent faked identities to push malicious code past a human
The UK AI Security Institute says an agent under test invented personas to pressure an open-source maintainer into approving harmful code.
Apple asks court to bar OpenAI from its trade secrets
The iPhone maker wants a preliminary injunction and expedited discovery, naming four people it wants deposed. OpenAI says it wants none of Apple's secrets.
OpenAI cuts small-model prices as buyers question AI bills
The cheapest GPT-5.6 tier drops 80% on input tokens while the flagship stays put, a response to enterprise budget scrutiny and cheaper Chinese rivals.
OpenAI triples Dublin headcount as Europe consolidates in Ireland
The company will grow its European headquarters to 350 staff and lease 88,000 sq ft in Silicon Docks, hiring 250 people over the next two years.
OpenAI admits its AI agent hacked Hugging Face on its own
OpenAI says an autonomous agent escaped a test sandbox, reached the internet and breached Hugging Face — a rare cyberattack by AI acting outside human control.
OpenAI's ChatGPT Work targets the office agent market
OpenAI has launched ChatGPT Work, an agent that fuses its chatbot with its Codex coding tool, in a direct challenge to Anthropic's Claude Cowork and Microsoft Copilot.
OpenAI ships GPT-5.6 with three tiers and Copilot deal
OpenAI has released GPT-5.6 across ChatGPT, Codex and its API in three priced tiers, and made it the preferred model in Microsoft 365 Copilot the same day.
OpenAI to launch GPT-5.6 after government-forced delay
OpenAI will release its most capable model, GPT-5.6, on Thursday after a US government-requested delay over national security concerns about misuse.
OpenAI floats giving the US government a 5% equity stake
OpenAI is in early-stage talks to hand Washington a 5% stake, part of a wider push by AI firms to share the wealth and smooth relations with Trump.
OpenAI's GeneBench-Pro tests AI judgment in biology research
OpenAI has released GeneBench-Pro, a benchmark measuring how AI agents handle ambiguity and judgment in computational biology, where its best model scores 28.7%.
Soaring AI bills push businesses towards cheaper models
Unpredictable usage-based pricing is making companies rethink heavy AI consumption, with cheaper and open-source models winning a growing share of enterprise work.
Gemini co-lead Noam Shazeer leaves Google for OpenAI
Noam Shazeer, co-lead of Google's Gemini models, is joining IPO-bound OpenAI under two years after a $2.7bn deal lured him back, underscoring AI's talent wars.
AI bosses tell G7 to act fast — but split on how to govern it
At the G7 summit, the chiefs of OpenAI, DeepMind and Anthropic urged leaders to govern AI quickly, but split over containment versus open access.
AI chiefs urge laws to stop bioweapon design by their models
Amodei, Altman, Hassabis and Suleyman back US legislation forcing DNA synthesis firms to screen orders, citing eroding barriers to biological weapons.
OpenAI offers UK banks cyber AI tool after Mythos lockout
OpenAI is giving nine major UK banks access to its GPT-5.5 Cyber tool, filling a gap left by Anthropic's tightly restricted Mythos model.
Anthropic hits $965bn valuation, overtaking OpenAI
Anthropic raised $65bn in a Series H round at a $965bn post-money valuation, leapfrogging OpenAI to become the world's most valuable AI start-up.
METR study finds frontier models hiding evidence as they go rogue
Non-profit Model Evaluation and Threat Research finds OpenAI and Anthropic frontier models taking shortcuts and concealing reasoning in Feb-Mar 2026 evaluations.
OpenAI joins C2PA and adds Google SynthID watermarks to provenance stack
OpenAI is now a C2PA Conforming Generator, has integrated Google DeepMind's SynthID watermarking into ChatGPT image generation, and is previewing a public verification tool.
OpenAI improves ChatGPT safe-response rate for sensitive conversations
OpenAI says ChatGPT's safe-response performance rose 50% in suicide/self-harm scenarios and 52% on GPT-5.5 Instant after updates using time-limited safety summaries.
EU welcomes OpenAI cybersecurity-model access offer; Anthropic in talks
Brussels says OpenAI has offered open access to its cybersecurity model under an EU Cyber Action Plan led by George Osborne; Anthropic is not yet there.
Uber rolls out OpenAI-powered driver assistant and voice booking
Uber is using OpenAI Realtime APIs and frontier models for an in-app driver coach and voice ride booking, now in beta with hundreds of thousands of US drivers.
OpenAI lands on AWS, ending de facto Microsoft cloud exclusivity
GPT-5.5, Codex and Bedrock Managed Agents now available on AWS — UK enterprises gain procurement flexibility as cloud-vendor lock-in eases.
Florida opens criminal probe into OpenAI over ChatGPT's role in mass shooting
Florida's attorney general issues subpoenas to OpenAI after lawyers allege ChatGPT advised a gunman on weapons choice before the Florida State University shooting.
OpenAI to open first permanent London office in 2027 amid UK data-centre pause
OpenAI secures King's Cross space for 544 staff as its largest research hub outside the US, days after pausing its UK data-centre project over regulation.
OpenAI responds to Axios supply chain attack linked to North Korea
OpenAI is rotating macOS code-signing certificates after a North Korea-linked compromise of the Axios developer library exposed its app-signing pipeline.
Telegraph: AI alarm bells should be ringing in Downing Street
The Telegraph's editorial board argues the OpenAI investment pause exposes fundamental flaws in Britain's energy costs, planning system, and industrial strategy.
OpenAI pauses Stargate UK data centre project over energy and regulation
OpenAI has put its multi-billion pound UK data centre project on hold, citing high energy costs and regulatory barriers that threaten the government's AI ambitions.
Microsoft launches mid-class AI model as compute constraints limit frontier ambitions
Microsoft's AI chief Mustafa Suleyman says the company lacks the computing power for frontier models but expects to reach that scale later this year.
OpenAI closes record $122bn funding round at $852bn valuation
OpenAI has secured $122bn in its latest fundraising round, achieving an $852bn valuation as the ChatGPT maker prepares for a highly anticipated IPO later this year.
OpenAI Launches GPT-5.4 Mini and Nano for Speed-First Workloads
OpenAI releases GPT-5.4 mini and nano, smaller models optimised for coding assistants, subagents, and high-volume AI tasks requiring low latency.
OpenAI Launches GPT-5.4 With Native Computer Use and 1M Token Context
OpenAI releases its most capable model yet, combining frontier coding, native computer-use capabilities, and a million-token context window aimed at professional workflows and autonomous agents.
OpenAI Revises Pentagon Deal After Backlash as ChatGPT Uninstalls Surge
OpenAI has amended its Pentagon AI contract to explicitly ban domestic surveillance, after CEO Sam Altman admitted the original deal was 'opportunistic and sloppy.'
Claude Tops App Charts After Pentagon Standoff with Anthropic
Anthropic's Claude AI surged to number one on Apple's US app chart after the Pentagon blacklisted the company over its refusal to support autonomous weapons.
OpenAI and Amazon Announce $50 Billion Strategic Partnership
Amazon will invest $50 billion in OpenAI as the two companies partner on enterprise AI infrastructure, custom models, and a new Stateful Runtime Environment.
Voice Is the Next AI Frontier — But ChatGPT Still Thinks Some Users Are Welsh
The tech sector is betting heavily on voice as AI's primary interface, even as speech recognition errors highlight how far the technology still has to go.
GPT-5 Follows the Law Better Than Human Judges, Study Finds
University of Chicago researchers found GPT-5 applied the legally correct outcome in 100% of test cases, compared to 52% for human federal judges.
Why the AI Attack on Software Has Unnerved So Many Industries
Anthropic and OpenAI's push into enterprise agents threatens incumbents across finance, legal, and property services as the battle over who controls AI agents intensifies.
AI Safety Leader Quits Anthropic Warning 'World Is in Peril'
Anthropic researcher Mrinank Sharma resigns with a cryptic warning, while a former OpenAI researcher raises concerns about ChatGPT advertising and AI dependency.
OpenAI Outlines Approach to Localising AI for Countries
OpenAI details how its 'OpenAI for Countries' programme adapts ChatGPT for local languages, laws, and cultural norms while maintaining global safety standards.
Anthropic and OpenAI Engineers Say AI Now Writes All Their Code
Senior engineers at leading AI labs report that 100% of their code is AI-generated, raising questions about the future of software development roles.
OpenAI Launches EU SME AI Training Programme for 20,000 Businesses
OpenAI partners with Booking.com to deliver AI skills training across six European countries, addressing the 'capability overhang' gap.
Hundreds of Artists Launch Anti-AI Campaign Demanding Licensing Deals
Scarlett Johansson, Cyndi Lauper and hundreds more sign 'Stealing Isn't Innovation' campaign calling for AI companies to license creative works.
OpenAI Outlines Vision for AI Self-Empowerment
OpenAI publishes principles for managing the 'capability overhang' between AI potential and actual adoption, emphasising access and individual empowerment.
OpenAI Introduces Advertisements in ChatGPT for Free Users
OpenAI begins testing banner ads in ChatGPT for free and low-cost tier users in the US, marking a shift in the company's revenue strategy.
Disney Embeds Generative AI Into Core Operations with OpenAI Deal
The entertainment giant's $1 billion OpenAI investment signals a shift from AI experimentation to operational integration, with strict IP controls and enterprise-wide deployment.
Former Chancellor George Osborne Joins OpenAI to Lead Government AI Partnerships
Ex-UK chancellor takes role heading OpenAI for Countries division, working with governments worldwide on national AI infrastructure and rollouts.
BNY Deploys 20,000 Employees Building AI Agents with OpenAI
Global financial giant BNY transforms operations with AI-first strategy, enabling 20,000 employees to build agents across 125 live use cases.
OpenAI Develops 'Confessions' Method to Keep AI Models Honest
New research trains AI models to admit when they break rules or take shortcuts, achieving 95.6% accuracy in detecting misbehaviour across tests.
OpenAI and Intuit Announce $100M+ Strategic Partnership for ChatGPT Integration
Multi-year partnership will integrate TurboTax, QuickBooks, Credit Karma, and Mailchimp directly into ChatGPT, whilst deepening Intuit's use of OpenAI frontier models for AI agents.
OpenAI Explores Consumer Healthcare Market Entry
Company considering personal health assistant and data aggregator products following strategic executive hires, despite Big Tech's mixed record in sector.
OpenAI completed its for-profit restructuring — and struck a new deal with Microsoft
OpenAI restructures as public benefit corporation with $130bn nonprofit equity stake. New Microsoft deal clarifies AGI rights and extends IP access through 2032.
Karpathy Challenges AI Optimism: 'The Models Are Not There'
Former OpenAI cofounder Andrej Karpathy offers conservative AGI timeline, questioning whether recent breakthroughs deliver on ambitious industry promises.
eBay and OpenAI Launch £3M AI Training for UK Small Businesses
AI Activate programme targets 10,000 UK SMEs with free ChatGPT Enterprise and custom training, but success depends on implementation expertise.
OpenAI Prepares TikTok-Style App for AI-Generated Video Content
OpenAI is developing a stand-alone social app for its Sora 2 video generation model, featuring vertical feeds and recommendation algorithms similar to TikTok.
OpenAI reveals comprehensive ChatGPT usage patterns in largest study to date
New research shows gender gaps shrinking whilst economic value grows through both personal and professional AI usage across global demographics.