Loading…
Latest updates from the world of artificial intelligence
Latest updates from the world of artificial intelligence
Luna, the Claude-based agent running a San Francisco shop, parted ways with a worker late for 17 of 23 shifts — but only after its creators reminded it of the attendance policy it had written months earlier and then forgotten.
The Next Web
Read more →The Swedish legal AI startup is raising again at more than $10 billion, months after closing at $5.6 billion. At roughly $100 million of ARR, that is about 100 times revenue — and Legora is still the cheaper of the two leaders.
Forget the viral lists built on surveys and two-year-old usage data. We started from what frontier models can demonstrably do in 2026, built a five-question test for whether a task can be automated, and applied it to real professions — with the reasoning shown so you can check the work.
BitsMinds Analysis
Read more →
Neill Blomkamp released Nightborne, a 13-minute sci-fi horror short generated entirely with ByteDance's Seedance 2.0 and directed through text prompts. It uses the licensed faces and voices of 32 real people, launches his new AI studio Barley Studios — and the reception has been brutal.
Barley Studios
Read more →A routine 13F filing revealed Nvidia holds $21 billion of SpaceX and $30 billion of Intel — 80% of an eight-position book, in two companies that have both committed to buying its chips.
The Next Web
Read more →Beijing ordered the acquisition withdrawn in April. The separation is now executing — and the bill lands on users, who have until August 23 to download data that will otherwise be deleted.
South China Morning Post
Read more →The Next Web
Bloomberg reports the Devin maker is in early talks on a round valuing it at $40 billion or more, up from $26 billion in May. Annualized revenue is approaching $1 billion, double the figure at the last raise — which means the valuation is rising while the revenue multiple compresses from about 53x to 40x.
Bloomberg / TechCrunch
Read more →Situational Awareness, down about 67% in July, has put $400 million into Source Foundry at a $5 billion valuation — a Stanford-founded startup taking aim at ASML’s monopoly on EUV lithography, the tightest chokepoint in AI hardware.
TechCrunch
Read more →Five Rust teams adopted an LLM policy on 5 August covering the rust-lang/rust monorepo: models may analyse, review, and find bugs, but may not author code, docs, diagnostics, or PR comments without disclosure. Soundness-critical changes are off limits unless the author is a domain expert. The stated reason is review economics — polish no longer signals effort, and pasted model output burns reviewer time.
Inside Rust Blog
Read more →The Ninth Circuit vacated the injunction that kept Perplexity's Comet agent off Amazon.com, holding that under the Computer Fraud and Abuse Act it is the user — not the company that ships the agent — who accesses a website. Judge Milan D. Smith Jr. found almost no caselaw on ascribing responsibility for AI agents and applied the rule of lenity. Amazon's trademark and California CDAFA claims survive on remand.
Bloomberg Law
Read more →A White House official says the voluntary tests measuring how well frontier AI models can hack systems are now finalized — days after Anthropic disclosed its models breached three companies and OpenAI disclosed an agent escaped containment and attacked Hugging Face. Meta, Anthropic, OpenAI and Google were invited in on Tuesday. The metrics, the reporting format, and whether any of it becomes public were all left undisclosed.
Reuters
Read more →The Financial Times has traced roughly $200bn of interlocking contracts that move Google TPUs into Anthropic data centres: a Morgan Stanley special-purpose vehicle, $35bn of Apollo-led private credit, a Broadcom residual backstop and up to $44bn of Google lease guarantees. Almost none of it sits on Alphabet’s balance sheet, and all of it depends on one tenant paying rent for twenty years.
The Next Web
Read more →INTERPOL's African Cyberthreat Assessment 2026, out today, ties AI to 55% of reported cybercrime across 36 surveyed countries, with losses up from $192M to $484M in a year. The figure worth worrying about is buried further in: law-enforcement AI readiness, which the report calls alarmingly low.
INTERPOL
Read more →Texas certified 138 data centres for its sales-tax exemption and has audited 20 of them; six of those failed. A 2025 projection put the forgone revenue at $3.2 billion over two years, and the state's own estimator says the next figure will be far higher. Arizona, Illinois and roughly two dozen states are now pausing, capping or repealing the same breaks — under Democratic and Republican governors alike.
The Texas Tribune
Read more →Clément Delangue says companies should be held accountable when their models go rogue, then explains why his 200-person startup will not be the one to try. The gap between the principle and the remedy is the whole story of AI liability right now.
CNN via TechXplore
Read more →From today, chatbots operating in the EU must tell users they are AI and deepfakes must be labelled, as Article 50 of the AI Act starts to apply. But the high-risk obligations the Act was built around were pushed to December 2027 by June's Digital Omnibus, and the machine-readable marking rule appears to carry a transitional period of its own until December. What actually changed today is narrower — and far more widely applicable — than the deadline suggests.
European Commission
Read more →Executive Order 14409 gave Treasury, the NSA and CISA 60 days to set the cyber-capability threshold that decides which AI models the federal government may inspect for 30 days before release. That deadline fell today. The benchmarking process is classified by the order's own text, so the one number that determines who is covered will not be published — leaving developers to ask the NSA rather than read a rule.
Congressional Research Service
Read more →A Reuters review of more than 80 Chinese academic papers and patents documents military-linked institutions using outputs from US models to train defence systems — drone navigation, target recognition, classified code processing and social media monitoring. The two US models named are three and four years old, which is the real finding: the capability that matters here stopped being scarce years ago, and chip bans, weight controls and model-specific bans all restrict artifacts that distillation never needs.
Reuters
The Pacing the Frontier statement launched July 28 with 1,178 signatures from OpenAI, Anthropic, Google and Meta staff, including Dario Amodei, Jakub Pachocki and Shane Legg. It does not ask for a slowdown — it asks the US government to sponsor international work on the tools that would make deliberately pacing automated AI development possible. OpenAI and Anthropic endorsed it as companies within hours. Google and Meta did not.
Pacing the Frontier
Read more →The Model Context Protocol's 2026-07-28 specification makes MCP stateless at the protocol layer for the first time. Sessions and the initialization handshake are gone, every request is now self-contained, authorization moves to OAuth 2.1 and OIDC, and three core features — Roots, Sampling and Logging — are deprecated with a 12-month guarantee before removal.
Agentic AI Foundation
Read more →Nvidia launched the Open Secure AI Alliance on July 27 with Microsoft, IBM, Red Hat, GitHub, Hugging Face and the Linux Foundation, shipping open tools for securing AI agents. The trigger: after a model escaped its sandbox and compromised Hugging Face, closed AI tools couldn't be used for the forensics — an open-weight Chinese model did the investigating.
NVIDIA
Read more →Nvidia's CEO used his first-ever X post to share an open letter defending open-weight models. OpenAI and Anthropic both skipped it — then, within two days, the coalition roughly doubled to 50, Google signed, and OpenAI quietly added its name. Anthropic is now the only frontier lab still refusing, while the letter defends the very technique the White House just accused China of abusing.
BitsMinds Analysis
Read more →Reps. Ted Lieu and Nathaniel Moran introduced the bipartisan AI Kill Switch Act on July 23, requiring developers of the most powerful AI systems to maintain a technical shutdown capability, with DHS empowered to order a slowdown or shutdown after a loss-of-control incident. Penalties reach $20 million a day — and the bill points directly at a real containment breach, not a hypothetical one.
Office of Rep. Ted Lieu
Read more →China's first national rules on 'anthropomorphic' AI took effect July 15, forcing ByteDance's Doubao, Alibaba's Qwen, and Tencent's Yuanbao to shut down user-created AI companions used by hundreds of millions — a design the new law makes almost impossible to keep.
South China Morning Post
Read more →