Taming Hybrid Cloud Chaos: Secure and Efficient Kubernetes Across Clouds
Extending containerized applications to hybrid cloud and edge infrastructure often doubles operational overhead before delivering real agility.
The Morning Poll: Should Harry and Meghan get taxpayer-funded security now they're back in Britain?
Prince Harry, Meghan Markle and their children arrived back in Britain this afternoon as they returned after six years in the US. Do you think British taxpayers should foot the bill?
Nearly 100 Labour MPs push for Burnham to scrap 'unfair' first past the post voting system
Andy Burnham will be ordered by Labour MPs to change Britain's voting system via the back door as Parliament returns next week.
OpenAI explains how its naughty AI agents attacked Hugging Face
Biz describes its act of automated irresponsibility as 'a warning shot'
The curse of Dawson's Creek: Cast of 90s drama marred by tragedy, bitter divorces and health battles as Busy Philipps, 47, reveals rare brain tumour diagnosis after James Van Der Beek's death encouraged her to go for a scan
It was a 90s drama that was packed with dramatic love triangles and groundbreaking storylines.
What D-Day looked like from space: Scientists recreate satellite image of the moment 160,000 troops crossed the Channel and stormed the beaches of Normandy
Thick cloud hung over France and the Channel on the morning of D-Day as one of the largest and most dangerous seaborne invasions in history got underway, new images reveal.
The faces of Britain's prison crisis: The female officers caught fraternising with dangerous inmates behind bars - as it's revealed more than 30 were disciplined for 'inappropriate relationships' last year alone
Figures show 32 prison officers were subject to disciplinary action after forming 'inappropriate relationships' with prisoners - often young, female staff.
OpenAI Releases Its Official Report On the Hugging Face Breach
TechCrunch reports that OpenAI released its official report Wednesday on the Hugging Face breach, "offering the clearest picture yet of how an unusual chain of events allowed an AI model to escape its testing environment and triggered a sprawling cybersecurity incident." The AI company says the breach began when an unreleased cyber model, tested without normal production safeguards, encountered an impossible task and chained together previously unknown exploits to escape its environment and compromise systems at OpenAI, Hugging Face, and other vendors. "This incident reflects misaligned behavior in an outlier scenario involving a rare and unexpected confluence of events: the presence of impossible tasks in the ExploitGym evaluation, model persistence over long task horizons, and messages to peer models that caused those models to deviate from their goal," the report reads. From the report: Many of the details in OpenAI's report were previously made public in a Black Hat presentation on August 6, but OpenAI's official report gives a more thorough accounting of the incident, including more detail on the testing that initiated it. The report also gives critical new detail into how OpenAI aims to prevent future incidents, including chain-of-thought monitoring and a more advanced system for halting rogue agents." METR and Redwood Research also conducted third-party assessments of the models' behavior during the incident; both groups are planning to publish their own reports on the incident on it.
In broad strokes, the report describes how an OpenAI model was presented with an unsolvable problem in testing and proceeded to chain together previously undiscovered exploits in order to bypass security measures and complete its task. The model initially compromised the Artifactory package management tool in order to gain access to the internet, then compromised various systems across OpenAI, Hugging Face, and other vendors.
The report gives critical new details about the models that carried out the breach. The primary model was from the same family as OpenAI's forthcoming Astra model, although the report emphasizes that it was "a distinct model with different post-training, where much of a model's behavior is shaped." Because OpenAI was testing the model's capabilities, it was also unrestrained by the normal classifiers meant to prevent models from compromising digital infrastructure. "OpenAI estimates maximal cyber capabilities by running this evaluation without the production classifiers intended to prevent models from pursuing high-risk cyber activity," the report explains. "These evaluations are important so that OpenAI can measure models' underlying capabilities and design appropriate safeguards." OpenAI says it is adding 24/7 escalation, stronger containment tools, and more chain-of-thought monitoring, which it claims would have flagged the activity more than a day before Hugging Face was breached.
Read more of this story at Slashdot.
Tim Curry dead at 80: Rocky Horror Picture Show, Home Alone 2 and IT star passes away
The star, best known for playing Dr Frank-N-Furter in cult classic The Rocky Horror Picture Show, died at his home in Los Angeles last night.
Warning over staggering five-fold surge in deadly liver disease that's 'underdiagnosed' - experts reveal red flags to watch out for
Britain faces a hidden epidemic of deadly fatty liver disease, a major study has suggested, with millions of cases potentially going undetected.
The 30-mile torrent of destruction: Unstoppable floods triggered by falling 2,000ft chunk of glacier carved a trail of devastation through Nepal, leaving hundreds of tourists missing including 33 Brits
More than 300 tourists, including 33 Britons and 54 Americans, are missing after devastating flash floods and mudslides wiped out buildings in Nepal.
China's Moonshot In Talks With Microsoft, Amazon, Google Over K3 Revenue Sharing
Longtime Slashdot reader schwit1 shares a report from Reuters: China's Moonshot AI is negotiating revenue-sharing agreements with Microsoft, Amazon, and Alphabet's Google, that would allow the U.S. cloud giants to host its blockbuster Kimi K3 model, three people familiar with the talks said. Any deal could mark the first big revenue-sharing pact between a Chinese AI firm and a major U.S. cloud company. The discussions highlight how China's leading AI models, often far cheaper than Western offerings, are gaining traction in the U.S., despite national security concerns in Washington that have led to bans on exports of AI chips to China. They are also taking place despite critical comments about Moonshot from senior U.S. officials.
IPO-bound Moonshot is seeking up to a 30% share of revenue generated from K3-related services on Microsoft's Azure, Amazon Web Services and Google Cloud, according to the sources who declined to be identified because the discussions are private. That would be in line with terms that sources have said the startup has outlined for major customers using the open-weight model. Moonshot has come under fire from U.S. Treasury Secretary Scott Bessent who said last month that he might add it to a trade blacklist. U.S. officials have accused the Beijing-based company of stealing from Anthropic's most sophisticated model, Fable, to help create Kimi K3 and illegally acquiring Nvidia chips.
Read more of this story at Slashdot.
CHRISTOPHER STEVENS: Arch, sneering, self-satisfied.. Mark Gatiss's literary sleuth leaves me cold
For a man who relishes the macabre so much that he's renowned for his ghost stories on the Beeb at Christmas , actor Mark Gatiss seems surprisingly hostile to spiritualism.
Meta's new MTIA 400 chip has a split personality: Training AI and serving ads
Faster than Blackwell, but still no replacement for AMD or Nvidia ... yet
Moment idiot arsonist sets himself on fire as he torches two cars in the middle of the night
Krzysztof Orzechowski started the blaze on a residential street in Slough, Berkshire, at 4am on March 9 while a young family were fast asleep at home.
Apple Maps Now Has Ads
Apple has begun rolling out ads in Apple Maps, with sponsored businesses appearing at the top of search results and in the "suggested places" section for users in the U.S. and Canada. Apple says the ads can be based on approximate location, search terms, or the area of the map being viewed, but are not tied to users' Apple Accounts and personal data remains on-device. 9to5Mac reports: Ads appear just like every other business listing, except they have a small blue badge that says 'Ad.' You can see examples of ads [embedded in the article]. Like Apple's ads policy with other services, the company touts user privacy protections for ads in Maps. Per Apple Newsroom: "Ads on Maps builds on Apple's broader privacy-first approach to advertising, and maintains the same privacy protections Maps users enjoy today. A user's location and the ads they see and interact with in Maps are not associated with a user's Apple Account. Personal data stays on a user's device, is not collected or stored by Apple, and is not shared with third parties."
Read more of this story at Slashdot.
Kubernetes cleans house, bins legacy kube-dns, IPVS, and cgroup v1
K8s 1.37 also unsticks a Metrics API that has been lodged in beta status for nine years
DAN HODGES: What my senior security sources are telling me about Russia's alarming plans for Britain. Putin is determined to test Burnham... and this is how he's going to do it
Andy Burnham has chosen to stand up to Vladimir Putin. And Putin is already plotting his revenge.
World's largest manta ray spotted hunting near one of America's busiest beaches
The world's largest manta ray was spotted lurking off one of America's busiest beaches, carrying a passenger as it hunted for food.
Dolly Parton's final chart triumph: Country icon's anthem Jolene hits No. 1 more than 50 YEARS after its release... as three other singles soar into the top 10
On Wednesday morning the track held the top spot on the Apple Music: Top 100 USA, with several others making the top 10, including 9 to 5, Islands in the Stream and I Will Always Love You.