Nicheloom

The opportunity tracker for new startups.

AI News

Real headlines from AI-dedicated outlets (TechCrunch AI, The Verge AI, OpenAI News, Hugging Face, and more) plus AI-relevant Hacker News stories, pulled every 6 hours — not another analysis, just the fastest real feed of what's happening.

The Sources/Vertical/Function filters above don't apply here — those filter launches, and a news headline isn't a launch. Use Source below to narrow by news outlet instead.

Source

Headline volume, last 14 days

Showing 7351–7400 of 9341 headlines.

GPT-5.2 derives a new result in theoretical physics Feb. 13, 2026, 11 a.m.

OpenAI News

A new preprint details how GPT-5.2 proposed a novel formula for a gluon amplitude in theoretical physics. The result was subsequently proved and formally verified by OpenAI alongside academic collaborators.

Introducing Lockdown Mode and Elevated Risk labels in ChatGPT Feb. 13, 2026, 10 a.m.

OpenAI News

ChatGPT has introduced Lockdown Mode and Elevated Risk labels to strengthen security for organizations. These features are designed to protect against threats such as prompt injection and AI-driven data exfiltration.

Beyond rate limits: scaling access to Codex and Sora Feb. 13, 2026, 9 a.m.

OpenAI News

OpenAI detailed the real-time access architecture created to power continuous access to its Sora and Codex models. The system integrates rate limits, usage tracking, and credits to scale beyond traditional access limitations.

Scaling social science research Feb. 13, 2026, 9 a.m.

OpenAI News

OpenAI has introduced GABRIEL, an open-source toolkit that leverages GPT to convert qualitative text and images into quantitative data. The tool is designed to help social scientists analyze and scale their research.

Introducing GPT-5.3-Codex-Spark Feb. 12, 2026, 10 a.m.

OpenAI News

OpenAI announced GPT-5.3-Codex-Spark, its first real-time coding model featuring a 128k context window and 15x faster generation speeds. The model is currently accessible as a research preview for ChatGPT Pro users.

Harness engineering: leveraging Codex in an agent-first world Feb. 11, 2026, 9 a.m.

OpenAI News

Technical staff member Ryan Lopopolo has authored an article on harness engineering and the utilization of Codex. The piece focuses on how to leverage the model within an agent-first software development environment.

Bringing ChatGPT to GenAI.mil Feb. 9, 2026, 11 a.m.

OpenAI News

OpenAI for Government has announced the deployment of a custom ChatGPT model on GenAI.mil. The initiative aims to provide U.S. defense teams with secure, safety-forward artificial intelligence capabilities.

GPT-5 lowers the cost of cell-free protein synthesis Feb. 5, 2026, 11 a.m.

OpenAI News

An autonomous laboratory pairing OpenAI's GPT-5 with Ginkgo Bioworks' cloud automation has reduced the cost of cell-free protein synthesis by 40%. The significant cost reduction was achieved through closed-loop experimentation.

Introducing Trusted Access for Cyber Feb. 5, 2026, 10 a.m.

OpenAI News

OpenAI has introduced Trusted Access for Cyber, a trust-based framework designed for cybersecurity use cases. The system expands access to advanced frontier cyber tools while implementing stronger protections to prevent misuse.

Introducing OpenAI Frontier Feb. 5, 2026, 6 a.m.

OpenAI News

OpenAI has unveiled OpenAI Frontier, an enterprise-focused platform for artificial intelligence agents. The service provides organizations with tools for building, deploying, and managing agents with shared context, user onboarding, permissions, and governance controls.

Vivo approves vlog camera project to rival DJI Pocket series Feb. 5, 2026, 2:59 a.m.

TechNode (China)

Chinese smartphone maker Vivo has greenlit an internal project to develop a dedicated vlog camera, targeting a planned launch in 2026. Designed to rival DJI's Pocket series, the portable device will feature a lightweight build and AI-powered stabilization tailored for content creators.

GPT-5.3-Codex System Card Feb. 5, 2026, midnight

OpenAI News

OpenAI has released the system card for GPT-5.3-Codex, describing it as its most capable agentic coding model so far. The system blends the specialized coding performance of GPT-5.2-Codex with the advanced reasoning and domain knowledge of GPT-5.2.

Introducing GPT-5.3-Codex Feb. 5, 2026, midnight

OpenAI News

OpenAI has introduced GPT-5.3-Codex, a Codex-native artificial intelligence agent. The model integrates frontier coding skills with general reasoning to assist with complex, long-horizon technical tasks.

Unlocking the Codex harness: how we built the App Server Feb. 4, 2026, 1 p.m.

OpenAI News

OpenAI has detailed the architecture of its Codex App Server, a bidirectional JSON-RPC API used to embed the Codex agent into workflows. The server interface provides capabilities including streaming progress updates, tool integration, user approvals, and code diffs.

LimX Dynamics raises $200 million in Series B to scale humanoid robotics Feb. 3, 2026, 3:48 a.m.

TechNode (China)

Shenzhen-based robotics startup LimX Dynamics has secured $200 million in a Series B funding round to scale its humanoid robot technology. Backed by investors including Lestone Capital, the company plans to use the capital to accelerate research and pursue global expansion.

The Sora feed philosophy Feb. 3, 2026, midnight

OpenAI News

The Sora feed philosophy focuses on fostering connections and inspiring creativity while maintaining a secure user experience. It integrates personalized recommendations, parental controls, and robust guardrails to ensure user safety.

Apple evaluates chip suppliers other than TSMC for first time in 12 years Feb. 2, 2026, 7:32 a.m.

TechNode (China)

Apple is considering shifting some of its low-end processor production away from TSMC, potentially ending a 12-year exclusive manufacturing partnership. The review comes as Nvidia surpasses Apple as TSMC's largest customer amid the current AI boom, securing a major portion of the foundry's advanced capacity.

Introducing the Codex app Feb. 2, 2026, midnight

OpenAI News

The newly introduced Codex app for macOS acts as an AI coding and software development command center. It is designed to handle parallel workflows, multiple agents, and long-running engineering tasks.

Operation “Date Bait”: AI-enabled scam targeting loveseekers Feb. 1, 2026, midnight

OpenAI News

Under Operation "Date Bait," OpenAI banned accounts originating likely from Cambodia that used AI to target Indonesian loveseekers. The perpetrators utilized the technology for translation, engagement, and scam outreach.

Romance scams: AI-enabled romance scam workflows Feb. 1, 2026, midnight

OpenAI News

OpenAI has banned accounts deploying artificial intelligence to power romance scam operations. The automated workflows were used for initial outreach, language translation, victim engagement, and promoting investment-fraud schemes.

Operation “Trolling Stone”: Russia-linked influence activity Feb. 1, 2026, midnight

OpenAI News

OpenAI banned accounts tied to "Trolling Stone," a previously undisclosed influence operation linked to Russia. The network utilized AI to generate comments regarding the arrest of a purported Russian cult leader in Argentina.

Operation “Fish Food”: Russia-origin content farm activity Feb. 1, 2026, midnight

OpenAI News

OpenAI banned accounts associated with the Rybar network, some of which likely originated in Russia, under an activity track designated Operation "Fish Food." The operation used AI tools to run multilingual influence efforts spanning various websites and social media platforms.

Alibaba’s T-Head unveils self-developed AI chip Zhenwu 810E Jan. 30, 2026, 5:42 a.m.

TechNode (China)

Alibaba's semiconductor division, T-Head, has launched the Zhenwu 810E, an in-house developed high-end AI chip. The processor features 96GB of HBM2e memory and 700GB/s of inter-chip bandwidth, delivering performance comparable to Nvidia's H20, with upgraded versions capable of exceeding the A100.

Project Genie: Experimenting with infinite, interactive worlds Jan. 29, 2026, 5:01 p.m.

Google DeepMind Blog

Google has made its Project Genie research prototype available to AI Ultra subscribers in the United States. The experimental platform enables users to create and explore infinite, interactive worlds.