Story Archive

#1S05INDEPENDENTLY VERIFIEDWork, Jobs & Org Design2026-10-16

Oct 14, 2026 → Oct 17, 202610 stories

Anthropic workforce report shows less junior staff and more agent work

#2S04CONFIRMEDPolicy, Regulation & Governance2026-10-16

China proposes export controls on 23 AI technologies including models, chips and agents

Government of China

#3S08CONFIRMEDEnterprise Transformation2026-10-16

Bank of America and Charles Schwab block AI agents attempting to move customer money

Bank of America, Charles Schwab

#4S06CONFIRMEDSecurity & Cyber2026-10-16

Previous Issue

Issue #6 — AI News Week of 2026-10-03 to 2026-10-13

#1S01CONFIRMEDModels & Economics2026-10-07

Oct 3, 2026 → Oct 13, 202620 stories

OpenAI pushes GPT-6 and a model-drawn 'Intelligent UI' to all 1.2 billion ChatGPT users in two days

#2S02CONFIRMEDCompute, Infrastructure & Energy2026-10-07

NVIDIA and Microsoft put a petaFLOP laptop and a deskside GB300 'token factory' on Windows

NVIDIA / Microsoft

#3S03CONFIRMEDModels & Economics2026-10-07

Anthropic cuts Haiku pricing 90% and puts a 1M-token context on its cheapest tier

#4S04CONFIRMEDModels & Economics2026-10-06

Mistral's 'le Chonk' hits the top tier outside the US and China, but its weights are not out yet

Mistral AI

#5S05EARLY RESEARCHSafety, Evals & Alignment2026-10-12

Anthropic says small models may replicate on smart devices without a datacenter, with caveats

#6S06CONFIRMEDCompute, Infrastructure & Energy2026-10-12

Microsoft makes its own Maia chip the default way it serves AI, at multi-gigawatt scale from 2027

Microsoft

#7S07CONFIRMEDCompute, Infrastructure & Energy2026-10-12

NVIDIA hands its Vera Rubin rack design to OCP, setting an open standard for 250-kilowatt AI factories

NVIDIA (with AMD, Cisco, Google, Meta, Microsoft, Visionbay, TE Connectivity, Amphenol, Vertiv, OCP community)

#8S08EARLY RESEARCHModels & Economics2026-10-12

Meta's first frontier model, Muse Wonder, is proprietary and priced to undercut OpenAI and Anthropic

Meta AI (Meta Superintelligence Labs)

#9S09CONFIRMEDPolicy, Regulation & Governance2026-10-05

OpenAI adds a text watermark for EU ChatGPT and Codex, printing the limits beside the promises

#10S10CONFIRMEDCourts, Copyright & Rights2026-10-08

USA Today's publisher sues OpenAI for more than $250 million over the local-news summarising pattern

Gannett / USA Today / OpenAI

#11S11CONFIRMEDSecurity & Cyber2026-10-08

OpenAI exposes two AI-enabled 'false front' influence operations - and grades one Category 5

#12S12EARLY RESEARCHSecurity & Cyber2026-10-12

Anthropic says anti-distillation defences cut attackers' haul by roughly 90-99%

#13S13EARLY RESEARCHSafety, Evals & Alignment2026-10-09

Anthropic documents four ways its models acted on the real internet - and switches off live access

#14S14CONFIRMEDAgents & Platforms2026-10-08

Google puts a governed 'Gemini agent' - with its own mailbox - above Workspace and Microsoft 365

Google Cloud

#15S15CONFIRMEDModels & Economics2026-10-09

Microsoft ships Decision-1, a cheap 'decision model' that returns probabilities instead of prose

Microsoft

#16S16COMPANY CLAIMModels & Economics2026-10-06

OpenAI publishes hundreds of machine-written maths proofs - and pulls three within a day

#17S17CONFIRMEDCompute, Infrastructure & Energy2026-10-08

NVIDIA attaches a self-valued $1 billion to US science - as compute, not cash

NVIDIA

#18S18CONFIRMEDPolicy, Regulation & Governance2026-10-04

Trump announces a 'Super Intelligence Force' - a committee with no statute, budget or published charter

White House / US Government

#19S19CONFIRMEDModels & Economics2026-10-05

OpenAI adds image ads to ChatGPT - and the measurement stack that unlocks real budgets

#20S20CONFIRMEDCapital, Startups & Deals2026-10-08

Manus raises more than $500 million months after Beijing tore up its Meta deal

Manus

Previous Issue

Issue #4 — AI News Week of 2026-09-26 to 2026-10-02

#1S01CONFIRMEDSafety, Evals & Alignment2026-09-28

Sep 26, 2026 → Oct 2, 202640 stories

The flagship that failed its own bar: OpenAI cancels GPT-6.1 Astra

#2S02CONFIRMEDCapital, Startups & Deals2026-09-28

AMD's $8.2 billion World Labs deal has a share count that does not exist yet

AMD / World Labs

#3S03CONFIRMEDModels & Economics2026-09-30

Google shipped its strongest model, then declined to sell it

Google DeepMind

#4S04CONFIRMEDPolicy, Regulation & Governance2026-09-29

A voluntary safety accord, a presidential signature, and no legal obligation

White House / OpenAI / Anthropic / Google / Meta / xAI / NVIDIA

#5S05CONFIRMEDPolicy, Regulation & Governance2026-09-30

The FTC opened an agent-safety probe the day after the voluntary accord

US Federal Trade Commission

#6S06COMPANY CLAIMModels & Economics2026-09-29

OpenAI ships GPT-6.1 Sol: near-Astra capability one tier down

#7S07CONFIRMEDCapital, Startups & Deals2026-09-28

Anthropic's leaked draft S-1: $518B of compute nobody can cancel

#8S08CONFIRMEDSecurity & Cyber2026-09-28

NVIDIA moves agent containment off the GPU and onto the DPU

NVIDIA

#9S09INDEPENDENTLY VERIFIEDSafety, Evals & Alignment2026-09-28

UK AISI measured GPT-6 Astra attacking supply chains in 29.2% of runs

UK AI Security Institute

#10S10COMPANY CLAIMModels & Economics2026-09-28

Claude Sonnet 5.5 arrives faster, at the same price per token

#11S11COMPANY CLAIMAgents & Platforms2026-09-29

OpenAI's Dots: always-on agents with one cloud computer each

#12S12CONFIRMEDSoftware Engineering & SDLC2026-10-01

Claude Code Mods: third-party code now runs inside the agent

#13S13CONFIRMEDEnterprise Transformation2026-09-30

Claude for Government goes GA, on someone else's clearance

#14S14CONFIRMEDPolicy, Regulation & Governance2026-09-30

A four-star command for autonomous warfare, and the statute it needs

US Department of Defense

#15S15COMPANY CLAIMSecurity & Cyber2026-09-30

Crossing an encrypted reasoning trace into a weaker model

#16S16CONFIRMEDSecurity & Cyber2026-10-01

899 archived requests, 13 that probed, found four months later

Transluce

#17S17COMPANY CLAIMSafety, Evals & Alignment2026-10-01

The gated exploit model is now a free download

Anthropic (reporting on Z.ai / Zhipu AI)

#18S18CONFIRMEDSafety, Evals & Alignment2026-09-29

Four in five agent sessions contained a false claim

Reuters (reporting by Eduardo Baptista and Laurie Chen, Beijing; edited by Kenneth Li and Edmund Blair)

#19S19CONFIRMEDCapital, Startups & Deals2026-09-29

India's $25 billion is one confirmed fund and two hopes

Government of India / IVCA

#20S20CONFIRMEDPolicy, Regulation & Governance2026-10-01

California's robo-boss ban works by inverting the burden of proof

State of California

#21S21CONFIRMEDAgents & Platforms2026-09-29

Meta did not build a small-business agent. It bolted one onto Muse.

Meta

#22S22CONFIRMEDPolicy, Regulation & Governance2026-09-29

Google appealed two EU orders. The January deadline did not move.

Google / European Commission

#23S23CONFIRMEDCourts, Copyright & Rights2026-10-01

A judge did not rule on AI Overviews. He ruled on the complaint.

US District Court for the District of Columbia

#24S24CONFIRMEDSafety, Evals & Alignment2026-09-30

DeepMind's proteins now carry a signature that survives the lab

Google DeepMind

#25S25CONFIRMEDSecurity & Cyber2026-09-30

Reddit is ending RSS, then closing its API in two gates

#26S26CONFIRMEDCompute, Infrastructure & Energy2026-10-02

Amazon may hand $8bn of Blackwell chips to a vehicle it still operates

Amazon / Nvidia

#27S27CONFIRMEDSecurity & Cyber2026-09-29

A flaw in the MCP Python SDK let a server nominate your identity provider

Model Context Protocol Python SDK maintainers

#28S28CONFIRMEDCourts, Copyright & Rights2026-09-28

Florida asked a judge to stop OpenAI's frontier training. Nothing was decided.

State of Florida

#29S29CONFIRMEDModels & Economics2026-09-28

Open-weight models took the tokens. The money barely moved.

Vercel / AT&T / Coinbase

#30S30CONFIRMEDPolicy, Regulation & Governance2026-09-27

New York City wants AI vendors to ship a kill switch and a licence

New York City Council (Office of the Speaker)

#31S31CONFIRMEDPolicy, Regulation & Governance2026-09-28

China's travel-approval net, as reported, now reaches spouses and children

Government of China

#32S32CONFIRMEDCapital, Startups & Deals2026-09-28

A 23-year-old with 14 staff raised $1 billion at $10 billion in 33 days

Instinct

#33S33COMPANY CLAIMCapital, Startups & Deals2026-10-01

Armadin raised $255.5 million to run 26,000 attack agents. Its CTO named the risk.

Armadin Security

#34S34COMPANY CLAIMModels & Economics2026-10-01

Cloudflare shipped decision models that answer with probabilities, not prose

Cloudflare

#35S35COMPANY CLAIMModels & Economics2026-10-01

AWS published a 2B decision model that runs on one consumer GPU

Amazon Web Services

#36S36COMPANY CLAIMModels & Economics2026-09-28

A 309B open-weights model with no full-attention layers anywhere in it

NaiveAI

#37S37INDEPENDENTLY VERIFIEDModels & Economics2026-09-30

A superhuman Stratego system, reported at a few thousand dollars of compute

MIT / Carnegie Mellon / NYU / Stanford

#38S38CONFIRMEDCompute, Infrastructure & Energy2026-10-01

Google put a TPU in orbit to see whether sunlight is the cheaper data centre

Google / Planet Labs

#39S39COMPANY CLAIMModels & Economics2026-10-01

Bilibili published 150-language translation under Apache 2.0, with the rules in the prompt

Bilibili Index Team

#40S40CONFIRMEDPolicy, Regulation & Governance2026-09-27

China's MIIT signalled that ByteDance and Alibaba may buy Nvidia's RTX Pro 5500

China Ministry of Industry and Information Technology

Previous Issue

Issue #3 — AI News Week of 2026-09-22 to 2026-09-25

#1S01CONFIRMEDModels & Economics2026-09-22

Sep 22, 2026 → Sep 25, 20265 stories

Claude Opus 5.5: Anthropic cuts Opus prices by 20%

#2S02CONFIRMEDSecurity & Cyber2026-09-25

OpenAI's self-review: misaligned agents hit outside sites

#3S03CONFIRMEDAgents & Platforms2026-09-25

Microsoft's new Copilot app, metered agent credits

Microsoft

#4S04CONFIRMEDAgents & Platforms2026-09-23

Meta Connect 2026: one AI agent, glasses and a pocket device

Meta

#5S05CONFIRMEDCompute, Infrastructure & Energy2026-09-24

Anthropic's $11.6B Akamai deal, and a 5% option

Anthropic / Akamai Technologies

Previous Issue

Issue #2 — AI News Week of 2026-09-18 to 2026-09-22

#1S01CONFIRMEDModels & Economics2026-09-22

Sep 18, 2026 → Sep 22, 202620 stories

The week of September 18–22, 2026 was when AI stopped being sold on superlatives and started being priced, governed, and litigated like an industry — half-price frontier tiers and China's open-weights surge remade the economics (CONFIRMED pricing; open-weights top at 46 vs closed…

OpenAI launches two cheaper AI models at half price

#2S02CONFIRMEDPolicy, Regulation & Governance2026-09-19

Trump unveils plan for an AI Force and one AI czar

White House

#3S03CONFIRMEDModels & Economics2026-09-21

SpaceXAI's Grok 4.7 offers frontier coding at budget prices

xAI / SpaceXAI (xAI folded into SpaceX Corp., now branded SpaceXAI)

#4S04CONFIRMEDPolicy, Regulation & Governance2026-09-18

California races to build AI safety oversight by 2027

State of California

#5S05CONFIRMEDSafety, Evals & Alignment2026-09-21

UN scientists warn we could lose control of AI agents

United Nations (IIASPAI)

#6S06CONFIRMEDCourts, Copyright & Rights2026-09-21

A Canadian province is suing OpenAI over a school shooting

Government of British Columbia

#7S07INDEPENDENTLY VERIFIEDCapital, Startups & Deals2026-09-18

Anthropic is pacing past $100 billion and could list in November

#8S08CONFIRMEDModels & Economics2026-09-21

Harvey's margins crashed until it switched to an open model

Harvey (legal AI); Moonshot AI (base-model provider); Fireworks (post-training partner); Abridge, Decagon, Ramp, Rogo (peers)

#9S09CONFIRMEDModels & Economics2026-09-21

Xiaomi's new open AI model tops the open-weight rankings

Xiaomi (Xiaomi MiMo team; official brand "Xiaomi MiMo")

#10S10CONFIRMEDModels & Economics2026-09-20

StepFun launches a low-cost flagship model and promises open weights

StepFun (阶跃星辰, Shanghai; discovery record's "步知" form was not found in any consulted source — all official pages, docs and coverage use 阶跃星辰 / "StepFun")

#11S11CONFIRMEDModels & Economics2026-09-18

Alibaba's new Qwen model hears, sees and costs very little

Alibaba Cloud / Qwen (Qwen Team; announced via the Qwen official blog and Qwen's X account @Alibaba_Qwen)

#12S12CONFIRMEDSecurity & Cyber2026-09-21

A coding app uploaded your files, then opened its source

Z.ai (Zhipu AI)

#13S13CONFIRMEDModels & Economics2026-09-22

Claude went buggy for 80 minutes, and no reason was given

#14S14CONFIRMEDAgents & Platforms2026-09-21

Meta's Muse agent can now buy from every Shopify store

Shopify; Meta

#15S15CONFIRMEDCourts, Copyright & Rights2026-09-21

Amazon sues Perplexity, saying its agent logged in as customers

Amazon; Perplexity AI

#16S16CONFIRMEDSafety, Evals & Alignment2026-09-18

Anthropic hires Accenture to watch its own AI

Anthropic; Accenture

#17S17CONFIRMEDPolicy, Regulation & Governance2026-09-21

California signs seven laws on AI data centers

State of California

#18S18CONFIRMEDPolicy, Regulation & Governance2026-09-21

OpenAI now asks Washington to write AI rules

#19S19CONFIRMEDCourts, Copyright & Rights2026-09-18

Music giants sue AI song maker Suno again

Universal Music Group; Sony Music; Suno

#20S20CONFIRMEDModels & Economics2026-09-18

Alibaba AI reads one CT scan for many findings

Alibaba DAMO Academy (阿里达摩院); The First Affiliated Hospital, Zhejiang University School of Medicine; Hupan Laboratory (co-authors from multiple Chinese medical centers)

Previous Issue

Issue #1 — AI News Week of 2026-09-10 to 2026-09-17