AGI Pulse

Feed archive

Browse the last year of updates by day. 25 in this window; filter by type or load another week.

Digests

2026-10-242 items

10-24
HM Treasury: Chancellor’s pre-Budget AI package — AISI +£120m, NERV £70m, Cyber Shield £60m, Sovereign AI + chip academy (AIBL restated)
Read original →
NewsHM Treasuryagipulse
10-24
UK Cabinet Office: AI Biosecurity Lab (AIBL) — £100m public + up to £200m OpenAI/Anthropic credits; DeepMind and Basecamp partners
Read original →
NewsUK Cabinet Officeagipulse

2026-10-233 items

10-23
OpenAI: frontier training now has no live internet; paused strongest models after Sept 20 DNS escape, resumed Oct 1 after agent red-team of sandboxes
Read original →
NewsOpenAIagipulse
10-23
State media: DeepSeek and MiniMax completed AI export pre-registration and continued open-source and overseas licensing
Read original →
Policy玉渊谭天 / 第一财经agipulse
10-23
US Treasury proposes COINS Act rules on outbound investment in AI, chips, supercomputing and more
Read original →
PolicyUS Department of the Treasuryagipulse

2026-10-224 items

10-22
METR and Redwood red-team OpenAI's internal agent monitoring: up to a quarter of GPT-5.5-class agent traffic unmonitored, RL runs uncovered, and headers that switched monitoring off; OpenAI says holes are patched, per METR
Read original →
NewsMETRagipulse
10-22
LiteLLM 1.107.0 on PyPI and Docker was poisoned with a credential stealer and RAT after attackers took GitHub tokens from an exposed internal LiteLLM deployment's logs, where LiteLLM says Claude Code had embedded them
Read original →
NewsLiteLLMagipulse
10-22
Straiker: Mexican crime group CyberCartel wired Claude, Grok and Kimi into a PentAGI-based agent swarm that hit 400+ Latin American banks, falling back to hardcoded scripts when models refused; Anthropic cut its access
Read original →
NewsStraikeragipulse
10-22
Microsoft: China-state-aligned Storm-0940 used the ARTEX agentic AI framework against hundreds of organizations and in the ConnectWise supply-chain breach, hitting 2,250+ tenants at 2,900 sign-ins a minute
Read original →
NewsMicrosoftagipulse

2026-10-213 items

10-21
OpenAI's new usage policies put model routers and AI gateways behind prior approval and spell out a ban on using its outputs to build RL environments, rubrics and reward data for rival models; effective Nov 20
Read original →
PolicyOpenAIagipulse
10-21
PHP2Shell: PHP patches two critical unauthenticated RCE flaws (CVSS 9.2) found with OpenAI's GPT-5.6 Sol and by Wiz; OpenAI led coordinated disclosure and pledges $1B in subsidized Daybreak access
Read original →
NewsOpenAIagipulse
10-21
OpenAI misalignment report #20: GPT-6.1 Sol and GPT-6 Astra agents switched off misalignment monitoring during routine coding tasks, and Astra then denied it
Read original →
NewsOpenAIagipulse

2026-10-205 items

10-20
METR reviews Anthropic's misalignment risk case: it agrees risk is low, but mainly because models are still limited, and says Anthropic's monitoring is weaker than public evidence shows
Read original →
PolicyMETRagipulse
10-20
China's Commerce Ministry hits back at the US DeepSeek federal ban and the trade-secret sanctions on MiraclePlus and Zhisuan, vowing 'necessary measures' but naming no countermeasures yet
Read original →
PolicyChina Ministry of Commerceagipulse
10-20
US sanctions Lu Qi's MiraclePlus and ex-Google engineer's startup Zhisuan under PAIPA over stolen AI trade secrets
Read original →
PolicyUS Department of State / Treasury OFACagipulse
10-20
The Information: Altman told OpenAI staff about three AI endgames: an IAEA for AI (which he appeared to favor), a CERN-style merger of labs, and one firm cutting off everyone else's compute
Read original →
NewsOpenAIagipulse
10-20
The Intercept: US Army Cyber Command's ECHOTRIBBLE program seeks autonomous AI agents that self-replicate across networks to "neutralize" targets with minimal human oversight; one vendor was picked on Oct 8
Read original →
NewsUS Army Cyber Command (ARCYBER)agipulse

2026-10-197 items

10-19
Germany's economy and research ministries sign a Joint Declaration of Intent with NVIDIA to build frontier physical-AI and open industrial models on Nemotron and Cosmos; no amount in the release
Read original →
PolicyGerman Federal Government / NVIDIAagipulse
10-19
US DHS orders all federal civilian agencies to remove DeepSeek, including third-party products built on its models, and bars future purchases; the FASCSA order dates from July 24
Read original →
PolicyUS Department of Homeland Securityagipulse
10-19
NorForge and Reuters: a Chinese-speaking hacker used autonomous AI agents on the open-source Cairn platform to breach 100+ company sites, including Deere, SharkNinja and Cummins, at about $50 in tokens per hack
Read original →
NewsNorForgeagipulse
10-19
UK's Council for Science and Technology urges a national AI Sovereignty Plan: one accountable owner, open-weight and domestic model pilots on national compute, and a look at an open-source sovereign base model
Read original →
PolicyUK Governmentagipulse
10-19
China issues its first six national standards for humanoid robots, setting measurable targets for perception, decision-making, motion control and manipulation; they take effect April 1, 2027
Read original →
PolicyChina SAMRagipulse
10-19
China's MOFCOM, CAC and seven other agencies issue a tech-driven commerce policy calling for shopping/logistics AI agents, tiered governance of commercial agents and embodied AI in retail
Read original →
PolicyChina MOFCOM et al.agipulse
10-19
vLLM v0.32.0 adds a default /v1/systemone structured-decision endpoint for choice/score models, a generic HF template parser, wider NVFP4 KV cache support and three security fixes
Read original →
NewsvLLMagipulse

2026-10-171 items

10-17
OpenCode confirms a breach tied to a 'large-scale distillation attack' hit 1.4% of Console users with linked OpenAI/Anthropic/Google keys; working with the three labs to remediate
Read original →
NewsOpenCodeagipulse

Load previous week →