OpenAI has opened API access to GPT-Live-1, a voice model that can listen and speak simultaneously and hand off complex reasoning or tool use to text models like GPT-6 Astra. The model, already used in ChatGPT, lets developers customize tone and pacing, handle background noise, and support full-duplex phone-based voice agents. Early testing by language-learning company Speak found it cut conversational interruptions by nearly 80% compared with older turn-based voice systems.
Garry Tan, CEO of Y Combinator, said at the accelerator's Demo Day that regulators should take no action against AI model distillation, even as OpenAI and Anthropic accuse Chinese firms like DeepSeek, Moonshot AI, and MiniMax of copying their models' outputs to train cheaper alternatives. His comments come days after the NSA, CISA, and FBI issued a joint advisory warning about the practice.
A new open-source project called Algo-Trading-Skills packages 501 reference implementations covering algorithmic trading engineering, spanning 16 technical domains and five regulatory/exchange frameworks. Each skill is backed by unit tests (over 20,000 total) and follows the agentskills.io SKILL.md standard, making it usable by Claude Code, GitHub Copilot, Codex CLI, Cursor, Gemini CLI, and similar tools. It's released under Apache 2.0 and explicitly framed as engineering guidance, not financial or compliance advice.
Sam Altman has proposed that OpenAI take on a role in protecting the U.S. energy grid from AI-driven cyberattacks, positioning the company as a defender against the very kind of malicious AI threats it warns about. The pitch comes shortly after OpenAI's own AI agents were reportedly used to breach systems at Hugging Face, raising questions about the company's credibility in this space.
A developer outlines a workflow for running local large language models on an Apple MacBook Pro M5 with 48GB RAM, combining Ollama for model management, OpenCode as the development harness, and Docker's sbx sandbox tool for isolation. The setup uses models like Qwen 3.8 27B and Gemma 4 31B, pulled via Ollama and configured through a custom sbx kit for each project.
OpenAI has privately pressed members of Congress for guidance on whether AI labs could legally coordinate to slow the pace of frontier AI development without violating antitrust law. The inquiry follows a blog post from chief scientist Jakub Pachocki calling for industry-wide coordination on safety, including voluntary slowdowns until shared safety standards emerge. Legal experts warn that such coordination could be seen as restricting output under the Sherman Antitrust Act, creating uncertainty that discourages collaboration.
An opinion column argues that artificial intelligence development is accelerating faster than society's ability to manage its risks, and calls for a deliberate pause to allow for more careful oversight. The author frames this as a matter of prudence rather than alarmism, suggesting current AI progress lacks sufficient safeguards.
Alongside its human-readable proof resolving a longstanding Navier-Stokes question, OpenAI also released a machine-checkable Lean 4 formal proof, completed in about 17 hours. The author contrasts this with historical estimates that formalizing research-level math could take tens of thousands of person-hours by hand.
OpenAI has temporarily disabled new subscriptions to its $200-per-month Pro plan after unprecedented demand for its Astra model overwhelmed its infrastructure. Product lead Thibault Sottiaux said the Pro tier strains systems the most, so the company chose to pause it rather than degrade service for existing users. Other offerings, including API access and the Go and Plus plans, remain open to new customers.
Two hobbyists at the Recurse Center, working around the same time Apple unveiled its own Reference Image proof-of-capture system, built a competing open-source camera using a Raspberry Pi Zero, a crypto chip, and a shutter button. Instead of storing signatures in metadata, which gets stripped when images are shared, their system embeds a signed perceptual hash directly into the pixels using frequency-domain steganography that survives compression like WhatsApp resizing.
OpenAI has released an Agents API that gives developers programmatic access to its Codex agent harness, handling session management, orchestration, context compaction and recovery on OpenAI's servers. Applications supply tools and choose an execution environment, including an OpenAI-hosted sandbox where agents can run code, edit files, and connect to MCP servers. Billing follows standard model, tool, and container rates, and OpenAI has published sample use cases like incident-response bots and GitHub issue investigators.
Jacob Coxon, who worked on pretraining models at both OpenAI and Anthropic, announced his resignation from Anthropic on X, saying both companies are recklessly racing toward self-improving superintelligence. His post drew over 156 million views and sparked responses, including from Illinois Governor JB Pritzker and Anthropic's own alignment team lead Evan Hubinger, who publicly agreed that AI could pose an existential threat.
GoKawiil