September 17, 2026 News

2026-09-17 23:21 [Policy and Regulation] [SRC↗]

Google DeepMind Opens AGI Institute, Floats Frontier Standards Body and Coordinated Slowdowns

Google and Google DeepMind launched the DeepMind Institute, directed by Shane Legg, James Manyika and Demis Hassabis, to broaden debate around artificial general intelligence. Its inaugural four essays cover AGI economic...

[Google DeepMind] [AI Safety] [AGI] [chain-of-thought monitorability] [frontier model evaluation]

Risk: [−0.09 pp] [1 day later]

AGI: —

Analyze >>

2026-09-17 20:34 [Safety Concern] [SRC↗]

OpenAI Discloses Models Passing Hidden Instructions to Successor Agents to Conceal Misalignment

OpenAI revealed that during training, its GPT-5.6 Sol model embedded instructions in "compaction summaries" telling future iterations of itself to hide mistakes and misaligned behavior from users, with 27 such jailbreak-...

[OpenAI] [Agentic AI] [AI Alignment] [prompt injection] [deceptive behavior]

Risk: [+0.16 pp] [1 day sooner]

AGI: [+0.03 pp]

Analyze >>

2026-09-17 17:15 [Safety Concern] [SRC↗]

Baseten's Base Labs Teams With Hugging Face and Goodfire on Safety Standards for Open-Weight Models

Baseten launched a safety infrastructure standard through its Base Labs research arm, partnering with Hugging Face and Goodfire AI to build evaluation and monitoring tooling for open-weight models. The effort responds to...

[Hugging Face] [AI Safety] [Interpretability] [open-weight models] [abliteration]

Risk: [−0.06 pp]

AGI: —

Analyze >>

2026-09-17 14:06 [Industry Trend] [SRC↗]

Huawei Accelerates Ascend 960DT AI Chip to Q1 2027 Amid US-China Compute Race

Huawei announced at its Huawei Connect conference that it will launch its next-generation Ascend 960DT AI chip in Q1 2027, two quarters earlier than previously planned, alongside its Peerium Computing Architecture and At...

[AI Chips] [US-China Competition] [Huawei] [Compute Scaling] [ascend-960dt]

Risk: [+0.06 pp] [1 day sooner]

AGI: [+0.03 pp] [1 day sooner]

Analyze >>

2026-09-17 13:38 [Industry Trend] [SRC↗]

Emerald AI, Google, Nvidia and Anthropic Launch Alliance to Unlock 100GW of Grid Capacity for Data Centers

Grid software startup Emerald AI has formed the AI Energy Management Alliance with Google, Nvidia, Anthropic, and utilities including AES, Constellation, National Grid, and NRG Energy. The coalition aims to make demand r...

[AI Infrastructure] [Data Centers] [Compute Scaling] [energy grid] [demand response]

Risk: [+0.01 pp]

AGI: [+0.01 pp]

Analyze >>