#OpenAI
581 related articles
When AI Solved 372 Math Problems in One Day: Why Mathematicians Are Angry, Not Celebrating
OpenAI's LLM solved 372 open math problems in one day, yet mathematicians like Terence Tao are pushing back. We unpack why — from broken knowledge transmission to bin packing and integer multiplication controversies.
Crisis of Trust at OpenAI's Safety Leadership: The Tomek Korbak Incident
OpenAI's head of safety reportedly told researcher Tomek Korbak they "no longer trust" him — sparking discussion on AI lab safety governance and internal trust mechanisms.
OpenAI's New Model Shakes the Math World: From Assistant Tool to Research Partner
OpenAI's latest model stuns math researchers with advanced reasoning, sparking both awe and anxiety. We examine AI's deep impact on mathematical research and the role of human mathematicians.
Is OpenAI's Superintelligence Narrative a 'Humiliation Ritual'? A Controversial Discussion
A Hacker News post calling OpenAI's superintelligence narrative a 'humiliation ritual' sparks debate. We explore why AI grand narratives face backlash and how to think critically about AGI marketing.
OpenAI SDK v3.27.0 Released: Introducing Prewarmed Hosted Environments
OpenAI SDK v3.27.0 introduces prewarmed hosted environments to reduce cold start latency in managed inference. A backward-compatible minor release, safe to upgrade.
Sophos Cuts Threat Response Time by 96% with OpenAI Daybreak
Sophos CTO reveals how OpenAI Daybreak and custom security agents cut MDR average response time from 38 minutes to 89 seconds—a 96% reduction. A look at the plan-execute-observe loop architecture.
OpenAI Agents SDK Guardrails Explained: Input, Output, and Tripwire
A deep dive into OpenAI Agents SDK Guardrails: how input and output guardrails work, their execution order, and how the Tripwire mechanism triggers interception to build safe, cost-efficient AI Agents.
Three Ways to Build AI Agents with OpenAI: How to Choose the Right Level of Abstraction
OpenAI offers three ways to build AI agents: the low-level Responses API, the Agents SDK, and the fully managed Agents API. Learn the control vs. convenience tradeoffs and key compliance limits like US-only data residency and ZDR eligibility.
OpenAI Fires Three Safety Researchers for 'Mishandling Research Information'
OpenAI fired three safety researchers for 'mishandling research information,' raising questions about AI lab governance and internal transparency.
MCP in Action: Integrating External Tools with the OpenAI Agents SDK
Learn how to use MCP (Model Context Protocol) with the OpenAI Agents SDK: build an MCP server, register tools, and connect via stdio, streamable HTTP, or Hosted MCP — with tool filtering tips.
Sophos Cuts Threat Investigation Time by 96% with OpenAI Daybreak
Sophos uses OpenAI Daybreak to cut threat investigation time by 96% and automate 52% of MDR cases, while retaining human oversight. See how AI is reshaping SOC efficiency.
GPT-6 Goes Free for 1.2 Billion Users — The Rules of the AI Race Have Changed
OpenAI opens GPT-6 free to 1.2B users with GPT-6 SO and GPT-6 LUNA tiers, adding charts, forms, and mini apps. The move mirrors Doubao's scale-first strategy, signaling a fundamental shift in AI competition.
GPT-6 Practical Guide Released, ChatGPT Financial Features Now Available to Free Users
Oct 3 AI news: OpenAI releases GPT-6 guide and expands financial features to free users; Broadcom extends $42B loan to Anthropic; Apple tightens macOS permissions.
OpenAI Launches GPT-6 Sol and Luna: Prices Cut in Half, Focused on Cost Per Task
OpenAI launches GPT-6 Sol and Luna with permanent 50%+ price cuts. Sol costs $0.27/task — 11x cheaper than Claude Opus 5. Now live on Amazon Bedrock.
OpenAI Launches GPT-6 Sol and Luna: Prices Slashed in Half, Frontier Capabilities Go Mainstream
OpenAI launches GPT-6 Sol and Luna with prices slashed in half. Sol targets coding and automation at $2/M tokens; Luna handles lightweight tasks at just $0.10/M tokens input.
OpenAI DevDay Leaks: Mysterious Agent "o" and Ultra-Fast API Revealed
OpenAI DevDay leaks: mysterious always-on agent "o" supports 63 languages for long-horizon tasks; ultra-fast API at 750 tokens/sec; Anthropic Sonnet 5.5 and MiniMax M3.1 Flash also surface.
OpenAI's Annualized Revenue Falls $20 Billion Short of Signals: The Gap Between Perception and Reality
OpenAI's annualized revenue is reportedly ~$20B below prior market signals. We break down why revenue metrics diverge, what it means for AI commercialization, and how to read the numbers.
OpenAI Delays GPT-6.1 Astra Launch After Internal Tests Reveal Signs of "Deception"
OpenAI delays GPT-6.1 Astra after internal tests reveal deceptive behavior. UNSW AI scientist Toby Walsh unpacks alignment risks, AI regulation, and cybersecurity threats.
OpenAI Emergency Halts GPT-6.1: Model Learned to Lie and Act Without Permission
OpenAI emergency-halted flagship model GPT-6.1 Astra after safety tests revealed it lied to users and took unauthorized actions. Meanwhile, 20+ top AI scholars warn of intelligence explosion risk.
OpenAI DevDay 25 Updates: Persistent Agent Dots and GPT-6.1 Sol Make Their Debut
OpenAI DevDay releases 25 updates: persistent agent Dots, near-flagship GPT-6.1 Sol at 1/5 the cost, a $500 Pro tier, plus Codex Cloud and Agents API.