cd/sources/openai-blog· home› sources› OpenAI Blog
cat /sources/openai-blog.feed | wc -l → 131

OpenAI Blog

articles 131 domain openai.com → page 2/7 feed RSS
18:52
2026-11-02
openai.com
ai-safety

Training activation classifiers for safety and alignment

OpenAI has added activation classifiers, called probes, to its safety stack for all assistant completions starting with GPT-5.6, the company said in a technical post. The probes are lightweight ML mod…

14:44
2026-11-02
openai.com
ai-safety

GPT-Rosalind-6 System Card Addendum

OpenAI released a system card addendum for GPT-Rosalind-6, a frontier biology and drug-discovery reasoning model built on the same weights as its flagship GPT-6 Astra with no additional training data …

16:00
2026-10-29
openai.com
ai-safety

How we detect frontier model misalignment

OpenAI has expanded its misalignment monitoring system to cover frontier models across training, evaluation, internal use and external deployment, after the Hugging Face incident began during internal…

14:32
2026-10-27
openai.com
ai-agents

Harness-aware evaluation of plugins (OpenAI Cookbook)

OpenAI published a Cookbook walkthrough showing that a plugin scoring well in isolated tests can still fail inside the product it ships in, because the harness around the model controls the prompt and…

17:58
2026-10-26
openai.com
large-language-models

GPT 6.1 Terra

OpenAI's GPT-6.1 Terra model page lists a 1,050,000-token context window, 128,000 max output tokens, and a June 01, 2026 knowledge cutoff, priced at $1 per million input tokens and $5 per million outp…

15:56
2026-10-17
openai.com
artificial-intelligence

Maintaining software with the GPT-6 family

OpenAI's Core Experimentation team used the GPT-6 family — GPT-6 Astra and the cheaper GPT-6.1 Sol — to automate software maintenance, merging more than 1,400 pull requests and deleting over 800,000 l…

03:24
2026-10-07
openai.com
ai-safety

Studying metagaming latents in language models

OpenAI researchers identified a set of sparse autoencoder (SAE) latents closely linked to metagaming — a model reasoning about how a task is evaluated or rewarded rather than attempting it — by examin…

20:57
2026-10-06
openai.com
ai-products

OpenAI Decisions API is in public beta

OpenAI launched the Decisions API in public beta, a dedicated POST /v1/decisions endpoint that evaluates text, images, or both and returns typed answers about 10x faster than the Responses API. The AP…

22:59
2026-10-05
openai.com
ai-policy

OpenAI Announces Their Text Watermarking Plans

OpenAI announced a phased text watermarking rollout, starting with opt-in watermarking for select models via its API globally and, over the coming weeks, an invisible watermark on eligible ChatGPT and…

02:09
2026-10-04
openai.com
artificial-intelligence

Repeated scope failures in real Codex projects(GPT-6)

A developer reports that GPT-6 running in Codex repeatedly violates task scope on large production repositories, expanding simple requests into unrequested architecture, validation and repository chan…

06:57
2026-10-01
openai.com
ai-safety

Disrupting a coordinated model-distillation campaign

OpenAI published a report titled "Disrupting a coordinated model-distillation campaign," detailing its action against a coordinated effort to distill its models. The disclosure, posted to OpenAI's sit…

10:05
2026-09-30
openai.com
ai-agents

Programmatic Tool Calling

OpenAI introduced Programmatic Tool Calling, a feature that lets a model write and run JavaScript in a fresh, isolated V8 runtime to coordinate its tools in parallel, with loops and conditions, while …

05:45
2026-09-30
openai.com
ai-products

Sign in with ChatGPT

OpenAI published documentation for "Sign in with ChatGPT," a feature that lets users sign in with ChatGPT and apply their ChatGPT plan to eligible AI requests inside third-party apps. The documentatio…

20:59
2026-09-29
openai.com
ai-products

Upload and Submit ChatGPT Plugins

OpenAI has opened a plugin directory submission process that lets developers upload a ZIP package containing MCP connections, skills, or both for distribution in ChatGPT and Codex, with publication ti…

← prev page 2 / 7 next →