AI News

onerail omnistar nvidia last mile delivery a signpost with three blank arrow boards

OneRail Uses Nvidia AI for Real-Time Last-Mile Delivery Optimisation — and the 10x Claim Has Three Different Shapes

OneRail announced OmniSTAR on 1 September 2026, a delivery decisioning platform built on Nvidia cuOpt and cuDF, and gave CNBC the exclusive. The headline was a tenfold speedup that turns a twenty-minute routing problem into a two-minute one. This is a documents-first read of that launch: the same speed claim published in three different shapes on the same day, the Apache 2.0 licence that makes the Nvidia engine free to every competitor, the word “first” tested against Nvidia’s own customer list of Domino’s, clicOH and Blue Yonder, and both disclosed deployments — a $39.4 billion foodservice distributor and an unnamed tyre distributor whose identity the vendor’s own newsroom narrows down. The product is real. The claims that travelled furthest are not the ones best supported by the record.

Read more
weather next 3 aixploria a satellite body with two flat wings

Discover Weather Next 3, Freshly Added to AIxploria — and the Rank That Puts It Behind Its Own Predecessor

AIxploria listed Weather Next 3 at 02:11 UTC on 6 September 2026 with a “new AI tool” badge, a Free tag, a 4.4 rating from seven votes and a corner line reading “#41 in Research & Science”. The predecessor it supersedes, WeatherNext 2, is badged #27 and sits exactly there in the directory’s own category listing. This is a documents-first read of that card: the vote count behind the stars, a 2.5-day gap between Google’s announcement and the listing, a claim-by-claim check against Google’s model guide, announcement post, access quick start and arXiv paper, and the two facts that decide whether the model is usable for you. Most of the card is accurate and one figure is scoped better than Google’s own blog. The rank and the omissions are the problem.

Read more
lyria 3 5 aixploria a lyria 3.5 record disc upright in slot holder

New AI Tool: Lyria 3.5 Has Just Landed on AIxploria — and Its Best Claim Is the One Google Never Made

AIxploria listed Lyria 3.5 at 03:10 UTC on 6 September 2026 with a “new AI tool” badge, a 4.5 rating from four votes and a corner line reading “#2 in Music”. The Music category page gives that second slot to Lyria 3, the predecessor, and does not carry the new model on page one or page two. This is a documents-first read of that listing: the vote count behind the stars, a 39-day gap between Google’s Flow Music release and the card, a claim-by-claim comparison of fifteen statements against Google’s model card, launch post and API price list, and the two claims that decide whether you should build on this model commercially. Ten check out exactly. The one carrying the card’s verdict is not supported by any Google Lyria page.

Read more
gpt 6 astra aixploria card a rubber stamp with round knob handle

GPT-6 Astra Is Now Available on AIxploria — and Its Card Claims a Rank the Directory Denies

AIxploria listed GPT-6 Astra at 03:35 UTC on 6 September 2026, three days after OpenAI shipped the model, with a gold verification badge, a 4.6 rating and a line reading “#1 in LLM models”. The site’s own LLM models category page does not contain the model on page one or page two. This is a documents-first read of that listing: the seven votes behind the star rating, a measured three-day propagation lag taken against two models we checked on the same pages earlier in the week, and a claim-by-claim comparison of twenty checkable statements against OpenAI’s launch post, model documentation, price list and system card. Sixteen check out exactly. Four do not, and two significant facts are missing.

Read more
roku ai slop channel fairground free a television set with blank recessed panel

Your Roku TV Offers AI Slop for Free — and the Contract Is the Real Story

Roku quietly added Fairground AI Creator TV, a round-the-clock channel of AI-generated films and AI-generated advertisements, at number 300 in its live TV guide. The reviews were savage and largely deserved. But the more interesting story sits underneath the jokes: a year-old company with $4 million in seed funding secured carriage on a platform Fox has agreed to buy for $22 billion, and it did so by solving distribution rather than generation. This is a documents-first look at what Roku actually added, who the named creators and studios really are, what Fairground’s published creator agreement says about the 50% revenue share, the 120-day payment window, the $100 payment floor and the $500 liability cap, and why a platform monetising roughly three cents per streamed hour finds a zero-marginal-cost channel structurally attractive.

Read more
gemini business custom mcp server connections a arch bridge span on two piers

Google Gemini Business Adds Support for Custom MCP Server Connections

Google has extended Gemini Enterprise – Business edition so a team administrator can register a custom MCP server connection and hand Gemini a callable interface to systems Google will never build a connector for. This is a documentation-first walkthrough of what that actually involves: the Connected apps path versus the Google Cloud data store path, the four hard requirements that disqualify most existing internal tooling on day one (StreamableHTTP transport, a publicly trusted TLS certificate, a public HTTPS endpoint and a publicly reachable identity provider), the OAuth field-by-field configuration, the 100-enabled-actions cap, the project-level quotas, the Pre-GA status caveat, and the security exposure a public MCP endpoint creates.

Read more
seattle times lawsuit openai microsoft copyright infringement a lectern with sloping top board

Two More News Organizations Sue OpenAI and Microsoft for Copyright Infringement

The Seattle Times Company and Newsday LLC sued OpenAI, eight affiliated OpenAI entities and Microsoft Corporation on 4 September 2026 in the Southern District of New York, docketed as No. 1:26-cv-07644. The 38-page complaint pleads seven counts — direct and vicarious copyright infringement, two DMCA claims over stripped copyright management information, and federal, Washington and New York trademark dilution — and asks the court to destroy the training datasets and models built on the plaintiffs’ journalism. This is a working read of the filing: the datasets named, the 88-word Boeing 737 MAX passage a model returned verbatim, the 47% referral traffic collapse behind the harm claim, the $300 million licensing market the plaintiffs were left out of, and where the case sits against the consolidated New York Times litigation that reached summary judgment the same day.

Read more
cyberkimi chrome v8 patch weaponized under a day a hourglass with two tapered bulbs

CyberKimi AI Claims It Weaponized a Chrome V8 Patch in Under a Day

CyberKimi, an unrestricted fork of Moonshot’s Kimi K3, is claimed to have turned a Chrome V8 security patch published on 2 September 2026 into a working sandbox-escape exploit in under 24 hours. Nobody has reproduced it, the demonstration video runs with the sandbox disabled, and the target build predates shipping Chrome Stable. This is a working read of the claimed exploit chain, the four things the evidence does not establish, Google’s separate and confirmed CVE-2026-85046 zero-day, the vendor’s published ExploitBench and CyberGym numbers, and the gap between those benchmarks and this claim.

Read more
GPT-6 Astra - openai gpt 6 astra critical cybersecurity threshold a boom barrier arm on upright post

OpenAI Launches GPT-6 Astra, Its First Model to Cross a Critical Cybersecurity Threshold

OpenAI shipped GPT-6 Astra on 3 September 2026 and rated it Critical for cybersecurity capability under its own Preparedness Framework, the first model of any lab to carry that tier. Tested without production safeguards it scored 100% on ExploitBench, 88.0% on SRE-Bench and 39.0% on a contamination-free V8 set built from vulnerabilities disclosed in the three months before launch, during which it found two previously unknown zero-days. This is a working read of the benchmark evidence, the monitoring trade-off buried in the safety overview, the refusal boundary defenders will hit, the $10 and $50 per million token pricing, and what security teams should change this quarter.

Read more
spytrend ad intelligence mcp server review a binoculars with two round barrels

SpyTrend Put Its 1.1 Billion-Ad Corpus Inside Claude, ChatGPT and Cursor

SpyTrend is a Facebook, Meta and TikTok ad intelligence platform built on one idea: research the operator, not the individual banner. It links ads by pixel, landing domain, Page ID and IP address into a single webmaster profile, claims 1.1 billion ads with 230 million added every month across 33 verticals, and since 2026 exposes the whole corpus to Claude, ChatGPT, Cursor, Codex and VS Code through an official MCP server on OAuth 2.1 with no API keys. This is a close reading of what it indexes, what the twenty tools return, what Pro’s 40,000 monthly tokens actually buy, what the vendor’s own test runs found, and the five places where its own pages contradict each other on refunds, verticals, free access and the price of a Meta row.

Read more
CHAT