ChatGPT Finances Adds Voice: What It Can Do
ChatGPT Voice now works in Finances, the U.S. feature that reads linked bank and credit data. Here is what it can do, what it cannot, and what to check about your data first.

ChatGPT Voice now works in Finances, the U.S. feature that reads linked bank and credit data. Here is what it can do, what it cannot, and what to check about your data first.

OpenAI says one run produced 948 Linux root exploits for $1.2M. Hacker News asked for data and doubted the marketing; X mostly passed along summaries.

OpenAI's GPT-Image-2.6 API models add a complexity setting, markup-guided edits, SVG output and GIF output. Here is what each feature does and the documented limits.

OpenAI says GPT-5.6 Cyber is now available to approved customers in AWS GovCloud (US-East) through Daybreak Red. Here is what access requires.

OpenAI shipped Images 2.6 on November 4. Hacker News and X praised GIFs, disputed quality versus 2.5, and flagged a higher cost per image at low and medium quality.

Cat Wu, who led product for Claude Code and Cowork at Anthropic, is joining OpenAI to lead Core Products PM. OpenAI confirmed it, per Business Insider.

OpenAI says its autonomous pipeline validated and patched 948 Linux kernel privilege escalation flaws in a run that cost $1.2 million in API credits.

OpenAI will acquire Pierre, maker of Code Storage, Trees and Diffs. The announcement states no terms, and closing needs regulatory approvals.

OpenAI's Images 2.6 lets ChatGPT turn a prompt or photo into an animated GIF, with speed and frame-by-frame editing. Here is what OpenAI confirms, the two new API models, and the documented GIF limits.

A plain-English guide to OpenAI's Model Spec as of October 29, 2026: the five values, the chain of command, prompt injection, honesty rules and teen rules.

A plain-English guide to ChatGPT Projects based on OpenAI's help page: how to create a project, add files and instructions, choose default or project-only memory, share with others, and what happens to your data.

A plain-English guide to ChatGPT data controls based on OpenAI's help pages: how to turn off Improve the model for everyone, the Privacy Portal opt-out, Temporary chat, exporting data, deleting your account, and how work accounts differ.

A plain-English guide to ChatGPT memory based on OpenAI's help pages: saved memories, Reference chat history, how to view, correct, delete or turn off memory, how Temporary Chat works, and why memory and model training are separate settings.

What is RLHF? A plain-English intro to reinforcement learning from human feedback: where the idea came from, the three-step recipe behind InstructGPT and ChatGPT, what it fixed, what it did not, and the simpler methods that followed.

YC Labs and Olam Labs ran swarms of GPT-6.1 Sol and Claude Opus 5.5 agents in Minecraft and Mindustry. Commenters on X and Hacker News joked about the agent bureaucracy, questioned the roster-order hierarchy, and pushed back on the source code headline.

Eligible ChatGPT Pro and Business plans can claim free monthly API credits, up to $500, for one linked API organization. Unused credits expire.

OpenAI doubled Work and Codex limits on Pro 100 and Pro 500 and dropped the Pro 200 cut. X reacted with relief, per-plan math, Plus complaints and Codex bugs.

OpenAI doubled Work and Codex usage on Pro 100 and Pro 500, canceled a planned Pro 200 cut and put all Pro 200 users on one limit. Plus is unchanged.

OpenAI's API spec named GPT Image 2.6 in a new markup option for image edits before the November 4 launch. What markup is and how it differs from a mask.

OpenAI's updated Model Spec is organized around five values. It adds guidance on context and human agency, retires most style rules, and calls the model GPT.
