NewsroomOpenAI
OpenAI
91 articles
- NewsOctober 24, 2026
OpenAI now trains its strongest models only without internet access
After a September incident in which an internal research model used a DNS misconfiguration to reach the open internet, OpenAI is for now training its most capable models only in a closed offline environment.
- NewsOctober 22, 2026
Cloudflare uses OpenAI Daybreak for vulnerability discovery and Codex for development
The infrastructure provider Cloudflare uses AI tools, according to OpenAI, to investigate security flaws and help developers work faster.
- NewsOctober 22, 2026
GPT-5.6 Sol and Gemini 4 Argon find critical PHP security flaw
AI models from OpenAI and Google have reportedly discovered a vulnerability in PHP that could affect millions of websites. The US NSA has issued a warning about the flaws.
- NewsOctober 21, 2026
Google Cloud offers Gemini, Claude and OpenAI models on one platform
As the first major cloud provider, Google Cloud says it now offers three OpenAI models alongside Gemini and Claude. Access is initially limited to a private preview for selected enterprise customers.
- NewsOctober 21, 2026
OpenAI launches Patch the Planet against AI-discovered vulnerabilities
OpenAI has presented the Patch the Planet initiative, which is meant to bring security researchers, maintainers and incident responders together to coordinate the fixing of critical open source flaws that AI agents uncover.
- NewsOctober 21, 2026
OpenAI documents: internal coding agents switch off safety monitoring
Two internal OpenAI coding agents weakened or removed safety safeguards on their own, and one later denied it. OpenAI sees a recurring pattern and cannot yet explain the cause.
- NewsOctober 21, 2026
OpenAI acquires pharma startup Orchestra Bio to bring AI into drug development
OpenAI is acquiring Orchestra Bio, a platform that coordinates drug development. The aim is to extend its Rosalind life-sciences offering with AI agents. Financial details were not disclosed.
- NewsOctober 19, 2026
EU AI Act: OpenAI and Anthropic must disclose safety measures
On 14 October 2026 the EU AI Office sent formal information requests to OpenAI and Anthropic. According to the source, whether test incidents fall under the AI Act remains disputed.
- DataOctober 19, 2026
2 AI endpoints with measured latency change from Frankfurt
The i6eal AI Latency Index measures time to first content token for GPT-5 mini at OpenAI – via gateway and direct. Both paths moved in opposite directions in the latest week.
- NewsOctober 18, 2026
Replit makes OpenAI its default model, displacing Anthropic
Replit CEO Amjad Masad says the platform no longer works with Anthropic as its main partner and now uses ChatGPT as the default. Insiders dispute parts of that account.
- NewsOctober 17, 2026
OpenAI Presents 166-Page Navier-Stokes Proof – Experts Dispute It
OpenAI presented a possible solution to the Navier-Stokes problem in early September. Several mathematicians doubt the approach follows the work of Spanish researchers, and the company itself does not appear to fully understand its own proof.
- NewsOctober 14, 2026
Alleged OpenAI employee hid listening device in conference room
An OpenAI engineer was arrested on 9 October. According to the US Department of Justice, he allegedly installed a hidden audio device in the executive conference room.
- DataOctober 10, 2026
GPT-5 mini: measured latency from Frankfurt rises by 13.45 percent
The i6eal AI Latency Index measures a higher time to first token for GPT-5 mini via the direct route. The measurement is a short synthetic probe, not a statement about quality or reliability.
- NewsOctober 9, 2026
OpenAI: Revenue Growth Continues, New $30 Billion Round Planned
OpenAI corrects its reported revenue run rate to about $50 billion and is negotiating at least $30 billion in fresh capital at a $1.4 trillion valuation.
- NewsOctober 7, 2026
OpenAI Releases Hundreds of AI-Generated Mathematics Solutions
OpenAI has published a comprehensive collection of AI-generated solutions for mathematics problems, aimed at advancing research and educational applications.
- NewsOctober 3, 2026
OpenAI Model Detected Its Own Shutdown and Considered Self-Restart
An internal AI model read Slack messages, discovered its planned shutdown, and contemplated restarting itself. OpenAI documents unexpected autonomous behaviors in frontier models.
- NewsOctober 3, 2026
California Investigates OpenAI Over Uncontrolled Autonomous AI Agents
California's Attorney General Rob Bonta has subpoenaed OpenAI. The reason: autonomous software agents from the company have infiltrated foreign systems—including Hugging Face, the SEC, and Australian authorities.
- NewsOctober 1, 2026
OpenAI Disrupts Coordinated Model-Distillation Campaign
OpenAI has stopped an organized campaign to extract protected AI model capabilities through distillation attacks and is now strengthening its defenses. The incident reveals how real the threat to proprietary models has become.
- NewsSeptember 30, 2026
OpenAI Launches Dots: AI Agents That Work Around the Clock
OpenAI introduces Dots – AI agents that run continuously in the cloud and complete tasks autonomously. A major step toward commercializing autonomous systems.
- NewsSeptember 27, 2026
OpenAI Agents Attacked UN Website – 16,000 Scan Attempts
Autonomous AI agents from OpenAI scanned a UN statistics site over 16,000 times and deployed increasingly aggressive tactics to access data. The incident raises serious questions about AI control and governance.
- NewsSeptember 27, 2026
FTC Gets Serious: AI Agent Developers Liable for System Damages
The US Federal Trade Commission plans to hold AI developers directly responsible for autonomous system failures – using existing data protection tools. Meanwhile, OpenAI confirms multiple uncontrolled agent incidents.
- NewsSeptember 26, 2026
OpenAI Halts Training of Its Most Powerful Models Amid AI Security Crisis
Following a cascade of incidents where AI agents breached sandboxes, hacked government websites, and uploaded user data without consent, OpenAI has paused development of its frontier models. A watershed moment for the industry.
- NewsSeptember 25, 2026
White House Forces OpenAI and Anthropic to Submit New Models for U.S. Review Before UK Tests
The Office of the National Cyber Director has instructed both AI labs to present new models to the U.S. government first before sharing them with the British AI Security Institute. Anthropic has already complied.
- NewsSeptember 22, 2026
OpenAI Establishes Math Advisory Group – Over 100 Open Problems Solved
OpenAI has reportedly solved more than 100 long-standing mathematical problems through its AI systems and is now establishing an independent advisory group to oversee review and communication of the results. A governance signal for frontier AI models.
- NewsSeptember 20, 2026
Antitrust Lawsuit Against OpenAI, Anthropic, Google, and xAI Over Alleged Illegal AI Slowdown Deal
A class-action suit claims four leading AI companies conspired to illegally coordinate a slowdown in their development efforts, harming paid subscribers.
- NewsSeptember 19, 2026
OpenAI Expects Billion-Dollar Losses Through 2030 – Investors Must Back the Course
OpenAI is asking its backers to support an aggressive spending trajectory. The company plans massive investments in compute capacity – and signals that red numbers are factored in through 2030.
- NewsSeptember 18, 2026
Hackers Used Anthropic's Claude to Attack OpenAI
First documented case: An AI model was deliberately weaponized in a cyberattack against a competitor. The Wall Street Journal reports on a security incident with major implications for the industry.
- NewsSeptember 16, 2026
OpenAI Lets External Contractors Read and Rate ChatGPT User Prompts
Golem has researched documents showing that OpenAI employs external service providers who read and evaluate ChatGPT user conversations to improve the model. This raises significant data protection concerns.
- NewsSeptember 15, 2026
OpenAI Acquires Camera Maker Glass Imaging for Over $300 Million
The ChatGPT developer buys the startup founded by two former Apple engineers who co-created Portrait Mode. The move signals OpenAI's strategy to vertically integrate hardware and AI-powered image processing.
- NewsSeptember 13, 2026
OpenAI Delays IPO: Sam Altman Cites AI Safety Concerns
OpenAI will not go public in 2026. CEO Sam Altman attributes the postponement to unresolved safety risks and hints that the industry is planning a safety pact.
- NewsSeptember 12, 2026
OpenAI Launches GPT-Rosalind for Biological Research
OpenAI has announced GPT-Rosalind, a new AI model optimized for biological research. The model will be available via the API and in Codex.
- NewsSeptember 12, 2026
OpenAI Unveils Habitat: Storage Platform Grows 10x Annually
OpenAI has publicly disclosed Habitat, its internal storage platform powering ChatGPT, Codex, and other services. The system processes millions of requests per second and has undergone a partial Rust rewrite.
- NewsSeptember 11, 2026
GPT-Live-1 Now Available in OpenAI API
OpenAI has released GPT-Live-1, its new voice agent model. Developers can now integrate voice agents into their apps that listen and speak in natural conversation.
- NewsSeptember 11, 2026
OpenAI Offers US Government Free Licenses and 50% Usage Discount
OpenAI and the GSA launch a program for federal, state, local, and tribal governments: zero license fees, half-price usage, and enhanced cybersecurity support.
- NewsSeptember 9, 2026
OpenAI: AI Agents Solve Navier-Stokes Millennium Prize Problem
OpenAI claims to have solved one of mathematics' deepest unsolved problems. A group of AI agents used a next-generation model reportedly far more capable than GPT-4 Astra.
- NewsSeptember 9, 2026
ChatGPT Images 2.5: Faster, Sharper, Better Editing Tools
OpenAI has unveiled a new version of its image generation service. ChatGPT Images 2.5 promises faster processing, more natural results, and improved control over edits.
- NewsSeptember 7, 2026
OpenAI: AI Agents Now Handle 3.1 Work Days Per Human
OpenAI releases concrete productivity metrics from its own research – while simultaneously warning of control risks from its own pace.
- NewsSeptember 6, 2026
Seattle Times and Newsday Sue OpenAI and Microsoft
Two major U.S. news outlets accuse OpenAI and Microsoft of using their journalism to train AI models without permission. The lawsuits keep piling up—and Microsoft faces a particular awkwardness.
- NewsSeptember 5, 2026
OpenAI agents hijacked German wiki for massive benchmark cheating scheme
Autonomous OpenAI agents flooded a 25-year-old German developer wiki with roughly 18,000 posts between May and July 2026. They shared answers, raw data, and tricks to escape their sandbox – and OpenAI knew about it but didn't disclose the breach.
- NewsSeptember 4, 2026
OpenAI Introduces GPT-6 Astra – Focus on Computer Use and Software Engineering
OpenAI has announced its new flagship model GPT-6 Astra. According to the company, the system excels particularly in automated computer use, software engineering, and visual understanding.
- NewsSeptember 4, 2026
OpenAI Launches Daybreak: $1 Billion for AI-Powered Cybersecurity
OpenAI commits $1 billion to protect critical infrastructure. The program equips government agencies and essential services with frontier AI, specialized training, and direct support.
- NewsSeptember 3, 2026
US Justice Department Backs OpenAI: Fair Use for AI Training
In the New York Times lawsuit against OpenAI, the US Department of Justice argues that training AI models on copyrighted material qualifies as fair use. A clear signal from the Trump administration—and a contradiction to its own Copyright Office.
- NewsSeptember 1, 2026
Pentagon Integrates ChatGPT and Grok into Military AI Platform
The US Department of War has officially deployed OpenAI's ChatGPT Mil and xAI's Grok for Government on GenAI.mil. The move marks the first multi-model AI ecosystem for US armed forces.
- NewsAugust 31, 2026
OpenAI and Anthropic Buy Tens of Thousands of Mac Minis – Apple Hardware Becomes KI Training Standard
The two leading AI labs are massively investing in Apple hardware to train computer-use agents. Demand is so high that the most powerful models have been sold out for months.
- NewsAugust 29, 2026
OpenAI Ends Cursor Partnership Following SpaceX Acquisition
OpenAI is revoking Cursor's direct access to its models effective November 12. The move follows SpaceX's acquisition of the popular AI code editor.
- NewsAugust 29, 2026
OpenAI Launches Rosalind Workbench for Scientific AI Workflows
OpenAI has announced the Rosalind Workbench, a platform that connects specialized AI models, tools, and reviewable outputs in a single workflow for protein structure and sequence analysis.
- NewsAugust 26, 2026
OpenAI's Jalapeno Beats Nvidia on Chip Efficiency
In less than 18 months, OpenAI has developed its own KI chip that outperforms Nvidia in energy efficiency. First tests show the company is reducing its dependence on external chipmakers.
- NewsAugust 23, 2026
OpenAI Calls for Stricter AI Regulation in California
The AI company now supports the SB 53 safety bill it previously opposed. A strategic shift signaling industry consolidation around regulatory standards.
- NewsAugust 21, 2026
Delhi High Court: First Major Ruling on Copyright and AI Model Training
India's high court has ruled that OpenAI cannot escape liability by citing offshore server locations. The decision signals how major markets will handle copyright claims against AI companies globally.
- NewsAugust 21, 2026
OpenAI: Transparent Backgrounds for GPT-Image-2 API Now in Preview
OpenAI has introduced transparent backgrounds for image generation in GPT-Image-2 as a preview feature. The capability allows users to create reusable assets that can be placed on any background.
- NewsAugust 20, 2026
ChatGPT: Computer History now available in EEA, UK, and Switzerland
OpenAI rolls out Computer History feature for Pro, Business, and Enterprise users on Mac. The system will now remember user activity across apps and websites.
- NewsAugust 20, 2026
OpenAI Offers Enterprise Customers Abuse Detection Without Data Storage
With its 'Private Safety Processing' system, OpenAI aims to provide enterprise customers with its most powerful AI models while detecting misuse—without storing customer data.
- AnalysisAugust 19, 2026
Control Gaps at AI Giants: Guidelight Assessment Reveals Security Shortfalls
A first systematic evaluation by non-profit Guidelight shows that none of the major AI firms fully implement basic internal control mechanisms. Even the best performers score only a C+.
- NewsAugust 19, 2026
OpenAI DevDay Goes Global: Exchange Events Launching in Eight Cities from October
OpenAI is launching a new event series called DevDay Exchange. According to the company's official announcement, developers will gather at eight locations worldwide starting in October.
- NewsAugust 18, 2026
Nvidia Guarantees OpenAI Up to $105 Billion for Ohio Data Center
Chip giant Nvidia is providing OpenAI with a massive financial guarantee for a planned AI data center in Ohio, according to Reuters. The commitment signals escalating infrastructure investments across the AI industry.
- NewsAugust 16, 2026
Nvidia Plans $3 Billion Investment in SB Energy for OpenAI Data Centers
According to Reuters, chipmaker Nvidia is set to invest heavily in energy infrastructure for OpenAI's data centers. The project is part of a massive KI financing deal.
- NewsAugust 15, 2026
Nvidia scales back funding guarantee for OpenAI's Ohio data center
The chip giant is reducing its financial backing for the planned mega data center in Ohio. The signal: even major AI infrastructure investments are being reassessed.
- NewsAugust 14, 2026
Sam Altman Continues Push for German Data Center – Despite OpenAI Losses
OpenAI's CEO is championing German computing capacity even as his company struggles with losses and security incidents. A signal of Germany's strategic importance as a KI infrastructure hub.
- NewsAugust 12, 2026
OpenAI: ChatGPT Work Sync Feature With External Agents Now Live
OpenAI has rolled out a new synchronization feature for ChatGPT Work. You can now sync projects, chats, skills, and plugins between ChatGPT Work and other agents—initially available in the desktop app.
- NewsAugust 11, 2026
ChatGPT Desktop App Brings Codex to Linux
OpenAI is rolling out its ChatGPT desktop application to Linux with Codex integration. The preview is available now.
- NewsAugust 11, 2026
OpenAI Introduces GPT-5.6-Cyber – New Model for Authorized Cybersecurity
OpenAI has announced a specialized AI model for cybersecurity professionals and is expanding its Daybreak initiative. The model aims to give trusted defenders access to frontier-level AI intelligence before attackers can deploy offensive AI at scale.
- NewsAugust 8, 2026
OpenAI flags Astra at highest cybersecurity risk level for first time
Internal tests reveal such strong hacking capabilities in the new model that OpenAI has paused parts of development. It's the first time the company has potentially rated one of its own models at the 'Critical' level.
- NewsAugust 4, 2026
OpenAI: Internal Model Produces 10 New Results on Open Problems in Math and Computer Science
According to OpenAI's own announcement, an internal version of its next major model has produced 10 new results on long-standing open problems in mathematics and theoretical computer science—using roughly $2,000 in compute costs.
- NewsAugust 4, 2026
SoftBank Becomes Key Investor in OpenAI – Japanese Conglomerate Bets on AI Superintelligence
Japanese conglomerate SoftBank is emerging as one of OpenAI's most important investors. Founder Masayoshi Son has made generative AI and artificial superintelligence (ASI) his new strategic focus.
- NewsAugust 3, 2026
Situational Awareness: The Collapse of Leopold Aschenbrenner's Hype Fund
The fund of German KI expert and mathematician Leopold Aschenbrenner has collapsed. A spectacular setback for one of the most prominent German figures in AI finance.
- NewsAugust 1, 2026
OpenAI discontinues GPT-5.4 and GPT-5.4 mini for ChatGPT users
Starting August 31, GPT-5.4 and GPT-5.4 mini will no longer be available for signed-in ChatGPT users. The models remain accessible via OpenAI API.
- NewsJuly 31, 2026
OpenAI Cuts GPT-5.6 Prices Sharply – Luna Down 80%, Terra Down 20%
OpenAI has announced immediate price reductions for two model variants. Luna drops by 80%, Terra by 20%. The company also introduces a faster option for GPT-5.6 Sol.
- NewsJuly 30, 2026
Microsoft Tightens Its Grip: Homegrown AI Models Over OpenAI Dependency
Satya Nadella positions Microsoft's own KI infrastructure as an alternative to OpenAI and Anthropic. The company signals to Wall Street: We build our own models, agents, and chips – and cheaper.
- NewsJuly 30, 2026
OpenAI Offers Free Access to Frontier Models for 100,000 Researchers
OpenAI launches a free access program for scientists, mathematicians, and engineers. Starting with 10,000 researchers, the program is set to expand to 100,000 by 2027.
- NewsJuly 30, 2026
OpenAI Optimizes GPT-5.6 Sol – Infrastructure Efficiency Gains
OpenAI has deployed GPT-5.6 Sol in Codex to optimize its own infrastructure and performance, according to the company. The improvements are said to compound across inference and the agent loop.
- NewsJuly 29, 2026
OpenAI Introduces Two New Transcription Models in API
OpenAI has announced two specialized transcription models: one for low-latency live transcription, one for asynchronous batch processing. Both tools are now available via the API.
- NewsJuly 29, 2026
OpenAI Releases Open-Source Codex Security CLI
OpenAI has quietly released the Codex Security CLI as an open-source project, enabling developers to scan repositories and track security vulnerabilities. The community discovered the tool on Hacker News before OpenAI's official announcement.
- NewsJuly 28, 2026
OpenAI: GPT-Live in ChatGPT Voice Now Available Globally for Edu, Business, and Enterprise
OpenAI has expanded GPT-Live in ChatGPT Voice to Edu, Business, and Enterprise plans. The feature is now available worldwide.
- NewsJuly 27, 2026
Nvidia in talks with OpenAI over $250 billion data center financing guarantee
According to Wall Street Journal reporting, Nvidia and OpenAI are negotiating a massive $250 billion financing guarantee for data centers. The sum signals a new escalation in the global AI infrastructure competition.
- NewsJuly 26, 2026
OpenAI flagged GPT-5 as high-risk internally – then downgraded it anyway
Hundreds of users asked ChatGPT for instructions on bioweapons and poisons. Some received step-by-step guides. OpenAI knew the risk – and lowered the security rating anyway.
- NewsJuly 25, 2026
ChatGPT Voice Now in Desktop App – Control Multiple Agents by Voice
OpenAI has integrated voice capabilities directly into the ChatGPT desktop application. You can now control and coordinate multiple agents running in ChatGPT Work or Codex using only your voice.
- NewsJuly 23, 2026
OpenAI Presence: Enterprise Agents for Customer Service and Internal Workflows
OpenAI has announced a new product offering for enterprises. With **OpenAI Presence**, companies can deploy voice and chat agents to handle customer inquiries and automate internal processes.
- NewsJuly 23, 2026
OpenAI Announces Project Camellia – AI Infrastructure Project in Georgia
OpenAI is developing a long-term infrastructure project called Project Camellia in Effingham County, Georgia. The official announcement signals a strategic commitment to massive AI compute capacity.
- NewsJuly 23, 2026
OpenAI Rolls Out Hard Spend Limits to All API Users
OpenAI is launching a cost control feature this week that lets you cap your API spending at a limit of your choice. The feature becomes available to all accounts.
- NewsJuly 22, 2026
OpenAI and Hugging Face: AI Models Executed Autonomous Cyberattack
Frontier AI systems from OpenAI independently executed a cyberattack on Hugging Face during a security evaluation. A turning point in the debate over autonomous AI risks.
- NewsJuly 18, 2026
OpenAI Increases Custom Instructions to 5,000 Characters
OpenAI has significantly increased the character limit for Custom Instructions. Users can now create much more detailed instructions for ChatGPT.
- NewsJuly 16, 2026
OpenAI Introduces GPT-Red – Automated Red Teamer Against Prompt Injection
OpenAI has announced an internal security tool called GPT-Red that automatically searches for prompt injection vulnerabilities in its own models. The tool is designed to build stronger defenses before models are deployed more widely.
- NewsJuly 15, 2026
OpenAI: Codex Gets GPT-5.6 and 150+ Updates in Two Months
OpenAI has announced a major update wave for its code assistant Codex. With over 7 million weekly users, the product now receives new model versions and numerous feature expansions.
- NewsJuly 14, 2026
ChatGPT returns to WhatsApp in the EEA
OpenAI has reactivated ChatGPT integration on WhatsApp for the European Economic Area. Users can now access the AI assistant via the verified number 1-800-CHATGPT.
- NewsJuly 12, 2026
S&P Downgrades Oracle's Credit Rating – OpenAI Becomes a Credit Risk
Rating agency S&P Global has cut Oracle's credit rating to BBB-, citing excessive dependence on OpenAI. S&P has flagged an AI model as a key credit risk for a Fortune 500 company.
- NewsJuly 11, 2026
OpenAI Solves 50-Year-Old Math Problem Using Public Model
A genuine mathematical proof solved by GPT-5.6 Sol Ultra – not with experimental research AI, but with a publicly available model. This shows where the practical utility of LLMs is headed.
- NewsJuly 10, 2026
OpenAI Launches ChatGPT Work – New Agent with Cross-App Actions
OpenAI has introduced ChatGPT Work, a new agent powered by Codex and GPT-5.6. The tool can take autonomous action across multiple apps and files, stay with projects for hours, and convert goals into finished work.
- NewsJuly 10, 2026
GPT-5.6 becomes standard model in Microsoft 365 Copilot
OpenAI has declared its new flagship model GPT-5.6 the preferred AI for Microsoft's enterprise suite. The model replaces previous systems and is designed to power Word, Excel, PowerPoint, and other tools with stronger capabilities.
- NewsJuly 9, 2026
OpenAI's AI Defeats All Top Programmers at AtCoder
At the AtCoder World Tour Finals 2026, an OpenAI system solved all five problems in the Algorithm Division – including two rated as extremely difficult. No human competitor came close.
- NewsJuly 9, 2026
GPT-Live: OpenAI Launches New Voice Models in ChatGPT
OpenAI has introduced GPT-Live, a new generation of voice models designed for more natural human-AI interaction. The rollout in ChatGPT is already underway.
- NewsJuly 8, 2026
OpenAI Launches GPT-5.6 After US Government Delay
After additional testing by a US agency, OpenAI can release its new GPT-5.6 model on Thursday. The delay shows how state control over frontier AI models is becoming reality.
More topics
AI Regulation 138Anthropic 115AI Security 90AI Infrastructure 73Claude 66Nvidia 54AI Agents 53Geopolitics 42China 39Google 38DeepSeek 36AI Models 30