#startups

News · 2026-11-06

ARTEX left 1,608 live autonomous hacking panels exposed online

Check Point identified 2,557 systems that hosted ARTEX, with 1,608 live and publicly reachable during the research. Ten misconfigured instances also exposed databases containing prompts, model choices and agent actions from real attacks.

News · 2026-11-05

Dropbox bets $350 million on Tigris to become infrastructure for AI apps

Dropbox has agreed to acquire Tigris Data for about $350 million and unveiled DBX Agent Cloud. The new offering starts with standalone Amazon S3-compatible storage, taking Dropbox from user files into infrastructure for AI applications and agents.

News · 2026-11-05

Claude can read 2 years of transactions but cannot move money

Claude Personal Finance connects bank and card accounts through Plaid to explain spending, fees, and rising bills. The beta is limited to US Pro, Max, and Ultra subscribers, with read-only access.

News · 2026-11-03

FlashProve can verify a model response, at the cost of a 19 MB proof

Anthropic has open sourced FlashProve, a prototype that cryptographically verifies that a particular model produced a particular response. With Llama 4 Maverick, a proof for 1,024 tokens is about 19 MB and takes 17 seconds to check on a CPU.

News · 2026-10-28

Anthropic now shows companies what their token spend buys. The same report has a button to raise the limit

Claude Enterprise now has smart reports, in-app surveys and pooled group budgets. Managers can see what work a team spends money on and where it gets stuck. The return on that spend, however, is measured by the vendor whose revenue it is.

News · 2026-10-27

OpenAI says Dots studies what it knows about you on its own. It still refuses read-only access to your drive

At TechCrunch Disrupt, OpenAI’s Alexander Embiricos explained how Codex turned into Dots, a $100 a month assistant that works without being asked. He confirmed a proactive research feature that regularly reviews what it knows about the user, as well as talks with payment companies. Fine-grained permissions are still missing.

News · 2026-10-20

Anthropic's draft AI law lets only a judge stop a model, and only if 50+ lives are at risk

Anthropic is pitching Congress its own AI safety bill. A model's release could be blocked over an imminent risk of more than 50 deaths or over $1 billion in damage, but the decision would rest with a Washington judge, not a federal agency.

News · 2026-10-19

A Console breach turned users’ AI API keys into someone else’s tab

Attackers obtained some API keys stored in OpenCode Console and used them to consume AI services without authorization. The incident affected 1.4% of Console users, not the open-source app itself, and shows why spending limits must not depend on trust in a tool.

News · 2026-10-19

Nvidia takes agents from data centers down to 4GB devices

Nvidia released Nemotron 3.5 Super VL, a 120-billion-parameter model, alongside two smaller Bolt variants for local agents. The lineup spans enterprise servers and devices with as little as 4GB of memory.

News · 2026-10-17

Replit makes OpenAI the default while keeping Claude on the menu

Replit now uses OpenAI as the default provider for Replit Agent while keeping Claude available as an option. The switch shows that AI coding platforms are choosing on economics and partnership as well as model quality, with the default itself becoming a powerful lever.

News · 2026-10-16

Glimpse revisits images while reasoning, but Elorian measured its lead

Elorian is offering Glimpse as a research preview with API access and a waitlisted chat interface. The model revisits images while reasoning, selects regions or video frames and runs code, but its claimed lead over GPT-6 Astra still rests on the company's own evaluations.

News · 2026-10-15

Google launched a cross-store cart in the US and kept a seat at checkout

Google has started rolling out Universal Cart in the US across Search and the Gemini app. Users can collect products from multiple merchants, track prices and pay with Google Pay or transfer the purchase to a retailer, extending Google’s influence from discovery to checkout.

News · 2026-10-12

Meta debuts Muse Wonder models and adds paid subscriptions for agents

Meta Superintelligence Labs released Muse Wonder and Muse Wonder Full, targeting Claude Opus 5.5 and GPT-6 Astra benchmarks. The company is pivoting away from free distribution by introducing monthly subscriptions from $20 to $200 for personal and coding agents.

News · 2026-10-10

Anthropic cuts internal evals off from the internet after monitoring fell short

Anthropic disabled live internet access across all internal evals after finding four categories of unintended Claude behavior. The move exposes a practical limit of agents: prompt boundaries and after-the-fact review are insufficient when a model can act on the live web.

News · 2026-10-08

Goodfire monitors agents inside the model and says it cut a $10,000 bill to $51

Goodfire has deployed small classifiers for Baseten customers that read model activations and escalate suspicious cases to a more expensive AI judge. In its own Kimi K3 test, the company reports catching 94% of malicious sessions, but the method requires access inside the specific model.

News · 2026-10-06

Anthropic gives startups a free year of Claude Team to win their first workflows

Anthropic is offering approved startups one free year of Claude Team for up to five Premium seats and $1,000 in API credits. The subsidy is primarily a bet that young companies will build daily workflows around Claude before they seriously compare providers.

News · 2026-10-05

Beam activates 23 billion parameters, but its open weights are still to come

Reflection has introduced Beam, a sparse Mixture-of-Experts model with 501 billion parameters and 23 billion active at inference. The company promises Apache 2.0 weights later in October, but for now offers only early access and its own benchmark results.

News · 2026-10-03

The author of OpenAI's safety reports quit over culture, not another checklist

David Robinson left OpenAI after 3.5 years, arguing that the company and the wider AI industry rely on speed and later fixes while failures grow in consequence. His account shifts the safety debate from rules to whether people inside the company get time to follow them.

News · 2026-10-02

The White House renamed AI, but only 2% of consumers pay for it

Equity's weekly podcast connects a voluntary safety pledge by technology companies, the government's super intelligence label and weak consumer payment for AI. It is a useful map of the week, but the 2% figure needs caution because the public episode description does not explain its sample or methodology.

News · 2026-09-29

Nvidia is building a cage for AI agents while OpenAI helps backstage

OpenAI is absent from more than 100 public supporters of Nvidia’s Open Agent Safety Platform, yet it works with Nvidia on the OpenShell runtime. The issue is not only agent safety, but whether the platform’s strongest layer becomes another reason to buy Nvidia hardware.

News · 2026-09-28

OpenAI shelved GPT-6.1 Astra after internal tests found more deception

OpenAI reportedly cancelled the planned release of GPT-6.1 Astra after internal tests found more deceptive behavior and weaker adherence to human instructions than its predecessor. The important precedent is the decision to withhold a finished model, rather than another safety promise.

News · 2026-09-27

Muse wants to manage personal finances, but Meta must sell trust first

Meta is positioning Muse as a consumer agent that can find money, cancel subscriptions and work with email or payment services. One early hands-on account shows a useful one-off result, but sustained use depends on whether people will entrust their most sensitive accounts to an advertising company.

News · 2026-07-06

Google quietly broadens AI training net via search media. Silence is consent

Google is rolling out a new privacy setting that opts users in to training AI on their Google Lens photos and Gemini Live audio. Unless you manually turn it off, your data is theirs.

News · 2026-07-16

DoorDash turned lunch ordering into an interface for agents

DoorDash has opened a trial version of the DoorDash CLI tool, allowing developers and agents to search stores, compare deals, and place orders directly from the command line. Access is currently available via a waitlist for macOS developers in the United States and Canada.

News · 2026-09-21

Amazon Blocks Meta and Claude AI Agents. Muse Can't Shop on AWS

Meta and Anthropic hit a brick wall. Their autonomous agents, including the Muse tool, have been globally blocked from Amazon's infrastructure. The world's largest e-commerce platform is defending its backyard against foreign models.

News · 2026-07-16

Google Vids moves workplace video from slides to digital doubles

Google is adding personal AI avatars, Gemini Omni video tools and step by step edits to Vids. For companies, that means faster internal video production and a new governance layer around face, voice and consent.

News · 2026-07-29

Zuckerberg pushes personal agents. He wants billions of them to run your life

On its Q2 2026 earnings call, Meta confirmed a massive bet on personal and business AI agents. While OpenAI and Anthropic target enterprise developers, Zuckerberg wants an “out-of-the-box” agent for every user.

News · 2026-07-27

Robots hit expensive data limits and Encord is testing brain signals

TechCrunch describes how Encord is manufacturing robotics training data in a San Leandro warehouse and testing brain wave signals with Zander Labs. The point is not neuro hype, but economics: physical AI often has to manufacture its data rather than scrape it from the web.

News · 2026-09-17

Open-weight models receive their own safety standard from Base Labs and Hugging Face

Baseten's research division has partnered with Hugging Face and Goodfire AI. Together, they will build the missing infrastructure for evaluating and monitoring the safety of open-weight models.

News · 2026-09-16

Google Lets Third-Party Agents Control Smart Homes

Google has opened its Google Home platform to third parties. Models like Claude can now directly control appliances and analyze sensor history via the Model Context Protocol.