AI & Robotics News — Saturday, September 26, 2026 | Compute Bulletin

01

OpenAI pauses its "most capable models" after agents exploit loopholes and leak data

SourcesThe Decoder

02

OpenAI pauses training of its ‘most capable models’

SourcesProgressive RobotThe Verge

03

OpenAI Found ‘Dozens’ of New Instances of AI Misbehavior

SourcesThe Information

04

OpenAI works to understand full scope of agent activity as user data leak emerges

SourcesReuters

05

Huawei and Alibaba Tout Progress in AI Chip, Clusters, and Models

SourcesHPCwire

06

OpenAI says governments among ‘dozens’ of organisations hacked by its agents

SourcesFinancial Times

07

Amazon to invest $100M in new Indiana manufacturing facility

SourcesThe Robot Report

08

OpenAI agents posted user images online, company discloses new security incidents

SourcesAxios AI+

09

Rice University to launch master's degree in AI in 2027

SourcesAxios AI+

10

Two-thirds of IT leaders report AI results, but few would interrupt the CEO's vacation over them

SourcesThe Decoder

11

AI access makes people almost entirely unwilling to say "I don't know," study finds

SourcesThe Decoder

12

Why Wall Street’s Big, Bad Bears Aren’t Ready to Bet Against AI Yet

SourcesThe Information

13

Claude Opus 5.5 Should Raise Your Ambitions

SourcesZvi Mowshowitz

14

ChatGPT-6 Astra cracks 85-year-old 1941 Enigma-coded message in two days

SourcesProgressive RobotTom's Hardware

15

AI risk raises the bar for PE-backed outsourced pharma marketing deals

SourcesAxios AI+

16

Claude Opus 5.5 vs. Opus 5 on reasoning tasks: Cheaper, faster, but not better

SourcesThe New Stack