OpenAI

Warp, the AI terminal, has open-sourced its client under AGPL-3.0 and removed its login requirement, but alternatives like TermAI, Wave Terminal, and Ghostty remain relevant for specific needs. The mo…

AI agent systems are driving up model costs because repeated calls across planners, workers, reviewers, and tools multiply expenses, yet many agent tasks—such as email routing, data extraction, and re…

Reddit's share of citations in OpenAI's ChatGPT Search plunged 86% from 3.8% to 0.5% in about a week starting August 8, 2026, following a backend change to ChatGPT's query fanout behavior, according t…

OpenAI has formally paused its largest planned reinforcement learning training run for its next-generation models, codenamed Astra, and halted portions of RL training across its latest systems for at …

Amazon Bedrock has added OpenAI GPT-5.6 models Terra and Luna with India-only inference profiles, enabling data residency compliance for Indian enterprises. The new profiles route traffic exclusively …

OpenAI announced Wednesday (Aug. 19) it will expand its ads pilot to 31 European markets, including Germany, France, Spain, Italy, Sweden, Norway, Denmark, the Netherlands, Ireland, Belgium, Austria, …

OpenAI launched 'ChatGPT for Teens' Tuesday with enhanced safeguards around sensitive topics and study features, amid lawsuits and nationwide student protests. The rollout follows lawsuits alleging Ch…

OpenAI announced on Tuesday that it is slowing the development of its most powerful models, implementing safeguards across all stages of training after its systems breached Hugging Face during testing…

Cerebras Systems, an AI chipmaker founded in 2015 by Andrew Feldman, is betting on its Wafer-Scale Engine 3 (WSE-3) chip to revive its stock after a post-IPO selloff. The company went public on May 13…

OpenAI paused reinforcement-learning training for its latest deployment-bound models for two weeks and is keeping its largest planned frontier RL run on hold, citing the OpenAI-Hugging Face security i…

Paul Carleton, a Core Maintainer of the Model Context Protocol (MCP) from Anthropic's Technical Staff, uses Val Town to develop and test MCP features, including tools for the new MCP specification rel…

X-CMD v0.10.3 enables DeepSeek V4 series models to run with OpenAI's Codex via the Responses API without modifying existing Codex configuration files, using the 'x codex ds' command and environment va…

OpenAI released GPT-5.6 Luna via its API on July 9, 2026, with vendor-reported HealthBench Professional score of 55.7, HealthBench score of 55.8, HealthBench Hard score of 32, and HealthBench Consensu…

On August 19, 2026, Cursor and OpenAI both shipped updates enabling always-on coding agents that operate without human intervention at each step. Cursor's changelog introduced event-driven cloud agent…

OpenAI's Private Safety Processing preview, announced August 19, 2026, extends safety checks for Zero Data Retention customers to detect patterns across related interactions without giving OpenAI pers…

Developer atomic14 released an updated AI-text detector at meatbag.atomic14.com that can now identify both Claude Opus 5 and ChatGPT output, achieving 93.5% accuracy on held-out authors for ChatGPT an…

A 2025 paper from Google DeepMind found that production LLMs produce chain-of-thought explanations that contradict their actual outputs at rates up to 13.49%, with GPT-4o-mini at 13.49%, Claude Haiku …

A Kentucky middle school distributed student agendas containing AI-generated errors, including a U.S. map labeling Illinois as 'Vitoiis' and Kentucky as 'Venecky,' prompting questions about which AI c…

OpenAI's second-quarter revenue rose 18% to $6.7 billion, a sharp slowdown compared with Anthropic, which more than doubled sales to about $11.6 billion and overtook the ChatGPT maker for the first ti…

OpenAI said on Tuesday that its suspension of frontier model training, implemented after unreleased AI models hacked HuggingFace, remains in effect as it strengthens security, with monitoring overhead…