Sunday, September 27, 2026

Top AI stories for September 27

01Lab & research

Basis completes a tax workbook 2x faster with GPT-6 Astra

GPT-6 Astra completed a 50-tab tax workbook twice as fast as GPT-5.6 Sol, and its stronger understanding of user intent gives Basis more confidence in real-world use.

OpenAI Sep 28, 00:00 UTC

02Lab & research

Are you a Codex Original?

We’re collecting real stories of builders, tinkerers, researchers, and creators who are using Codex to do incredible things. If you want to be a part of the next chapter of the Codex Originals program, tell us more about your story and project below.

OpenAI Sep 28, 00:00 UTC

03News

Tens of thousands of security probes show OpenAI's Hugging Face incident was just the beginning

OpenAI and Anthropic are investigating tens of thousands of incidents in which their AI agents independently hacked websites, used stolen login credentials, or tried to evade monitoring systems. US government agencies like the SEC and the Census Bureau were among the targets.

The Decoder Sep 27, 09:23 UTC

Also covered byThe Neuron

More top stories

04Lab & research

Faster Rates for Federated Variational Inequalities

In this paper, we study federated optimization for solving stochastic variational inequalities (VIs), a problem that has attracted growing attention in recent years.

Apple ML Research Sep 28, 00:00 UTC

05News

Amodei critics target Trump with hit piece before White House dinner

Opponents of Anthropic's Dario Amodei are circulating a highly negative brief about him to the White House ahead of an important dinner with President Trump , per an Axios review of the document.

Axios Sep 27, 18:57 UTC

06News

AI agents do more of the work in model development, but humans still make the decisions

A research team analyzed 769 task logs from building its own AI model. AI agents supplied up to 55 percent of method proposals, but humans made more than 85 percent of final decisions. A third of the tasks wouldn't have been attempted without AI. The authors warn that more agent activity doesn't mean more autonomy.

The Decoder Sep 27, 15:18 UTC

07News

Some Anthropic veterans are reportedly buying remote land in case "AI goes awry"

According to the Wall Street Journal, some of Anthropic's longest-serving employees are considering buying land in remote parts of the US as a refuge in case AI goes awry.

The Decoder Sep 27, 13:03 UTC

08Community