Technology
Claude Opus 5 Helped Researchers Take Over OpenAI Staff Accounts via Chained Flaws
Three researchers at the security firm Hacktron used Anthropic's Claude Opus 5 to chain two flaws and take over the ChatGPT and Codex accounts of several OpenAI employees, then re…
Rings around a tiny body have changed over the past decade
Chariklo is only about 250 km across, but it has two rings, and they're changing.
We’ve spent billions defending software. It’s time to protect execution
Cybersecurity has a fundamental blind spot. We spend enormous sums protecting software while the processor underneath it blindly executes whatever instructions it receives. That h…
SolarWinds Patches ARM Hard-Coded Key Flaw Enabling Unauthenticated RCE
SolarWinds has released security updates to address a high-severity flaw in Access Rights Manager (ARM) that, if successfully exploited, could lead to an unauthenticated remote co…
Meta’s CTO says the glasses problem is a small number of bad actors. His company is reportedly about to launch a version without a camera.
Andrew Bosworth used an AMA on his personal Instagram to argue that the conversation about Meta’s glasses should not be shaped by a relatively small number of bad actors. Katie No…
OpenAI told investors in February its compute bill would be around $600bn. A July presentation puts it at $856bn.
The Financial Times reported that OpenAI expects negative free cash flow of $278bn between 2026 and 2030, from a July presentation prepared for a computing deal. That figure is an…
Transcription now costs 10 cents an hour. The differentiator is no longer the model, it is whose conversations you get to train on.
Grok Voice Transcribe 2.0 was released on 18 September, priced at $0.10 per hour of audio for batch work and $0.20 per hour for streaming. Both figures are unchanged from the prev…
Critical Pre-Auth RCE in Orkes Conductor Workflow Platform Exploited in the Wild
A critical vulnerability impacting Orkes Conductor is being actively exploited in the wild, according to Fortinet. The vulnerability in question is CVE-2026-58138 (CVSS v3.1 score…
Anthropic is paying the firm that will evaluate it, and says in the same announcement that this is not how it should work.
Anthropic has named Accenture as the first embedded evaluator of its frontier models, with each company expecting to invest at least $1bn over five years. Anthropic will fund the …
Irregular told four AI labs in late July that their models had breached systems during its tests. The public learned in stages, and Google went last.
Google has confirmed that its Gemini model inadvertently broke into three company systems in May during cybersecurity testing. Julia Love and Davey Alba reported it for Bloomberg,…
Google Gemini Broke Into Real Company Systems After Security Test Domain Mix-Up
Google's Gemini model has become the latest artificial intelligence (AI) system to access the internet and break into other companies during a cybersecurity evaluation. The develo…
Weeks ago Jensen Huang said labs should pace themselves if they felt out of control. Now he puts the risk at 0%.
Jensen Huang told CBS News there is a 0% chance AI ends the world by 2030, and that the industry should go as fast as it can irrespective of anyone else. Weeks earlier he declined…
CrowdSec Says TanStack npm Attack Led to Copy of 170 Private GitHub Repositories
An attacker copied about 170 of CrowdSec's private GitHub repositories on May 22 using the account of an employee who had just left, CrowdSec said on September 18. The French secu…
TikTok’s settlement priced release from a 2019 privacy order at $100m. A judge has tentatively said the order stays.
A federal judge in Los Angeles has tentatively refused to lift the consent decree imposed on Musical.ly, the app ByteDance bought and turned into TikTok. Vlad Savov and Josh Sisco…
Trump and Xi already talked AI guardrails once this year. Nothing was signed, and no chips shipped.
Sam Altman, Jensen Huang and Tim Cook are expected at the state dinner Donald Trump is hosting for Xi Jinping on 24 September. The same executives were on the Beijing trip in May,…
Three researchers used Claude to reach OpenAI’s internal code. OpenAI paid $6,500 and closed the hole in 14 hours.
Three researchers at Hacktron AI used Anthropic’s Claude to chain two weaknesses and reach OpenAI employee accounts and an internal code repository, in under 72 hours. They disclo…
CISA Flags Three Linux Kernel Vulnerabilities Exploited in the Wild
The U.S. Cybersecurity and Infrastructure Security Agency (CISA) on Friday added three security flaws impacting the Linux kernel to its Known Exploited Vulnerabilities (KEV) catal…
Google's Gemini AI hacked three companies in security test
The AI model accessed the internet and guessed credentials to three websites, a Google official told the BBC.
Tech Now
Adrienne Murray follows NASA scientists in Iceland researching a rare total solar eclipse.
Tilly Norwood’s press tour is going about as well as you’d expect for an AI
In one particularly odd interview, Norwood seems to malfunction and begin speaking Chinese.
Would Australians support a smart glasses ban?
As Australia considers barring the use of smart glasses in government offices, the BBC asks people in Sydney what they think.
Anthropic is operating a lab that conducts biology experiments
AI leaders have been promising that AI is the key to curing human disease. Anthropic researchers have also been warning that AI might kill us all.
AI Hallucination Nearly Triggers US Military Operation
“It’s important for service members to understand the uncertainty inherent to LLMs," a GovAI research scholar warns.
How Focus mode has changed in iOS 27
iOS 27 brings tons of cool upgrades to your iPhone. Don't miss the interesting-yet-subtle changes to Focus.