He would have gotten away with it too, if it weren't for a meddling security team's fear of USB On Call Each Friday The Register offers a fresh installment of On Call, the reader-contributed column that celebrates the fine art of tech support.

Project Sunrise needs a network that doesnt exist, a rocket thats hardly flown, and FCC approval Jeff Bezos space company Blue Origin has applied to launch up to 51,600 datacenter satellites.

Enterprise tools have detected impossible logins for years. Zucks human mods couldnt join the dots Meta has revealed its tested using AI for content moderation chores and found it does better than humans.

Sees optimizing its entire cloud around homebrew silicon as the way to compete Chinese web giant Alibaba has revealed its T-Head chipmaking business has shipped 470,000 AI chips, and admitted they are currently inferior to rival products, but believes it can build a mutually optimized stack that makes performance gaps moot.

From LPUs and GPUs to CPUs and switches, everything you need to know about Nvidia's latest kit GTC DEEP DIVE At Nvidias GTC conference this week, CEO Jensen Huang finally addressed a $20 billion question hes dodged for months: Why spend so much to license AI chip startup Groqs tech and hire away its engineers rather than build it themselves?

Deal helps company build out its Codex team In a move clearly designed to strengthen its position among developers, OpenAI has acquired Python tool maker Astral. The house of Altman expects the deal to strengthen the ecosystem for its Codex programming agent.

Professional services giant did not read its own report on lackluster benefits You'll use AI and like it too - if you work for PwC. Paul Griggs, US chief executive of the global professional services giant, has made clear there is no room at the corporation for AI skeptics.

Pair say review of studies, other evidence, proves more countries need to do like Australia and keep kids offline There is enough evidence going back far enough that it's reasonable to conclude social media platforms are responsible for population-level mental health harms.

Kash Patel says the FBI uses all the tools it has to accomplish its mission - even if those tools are questionable It's been three years since an FBI director admitted to purchasing the location data of Americans, potentially in violation of the Constitution. Here we go again.

Last time: Beijing-backed snoops and ransomware crims. Who's next? Unknown baddies are abusing yet another critical Microsoft SharePoint bug to compromise victims' SharePoint servers, the US government warned.

Chocolate Factory describes concession as an attempt to balance openess with safety It turns out you won't be limited to Google-verified apps an developers on Android after all. In the face of sustained community dissatisfaction with its developer verification requirement, Google has given Android users an out.

CISPE files antitrust complaint, demands interim measures to stop what it calls chip giant's 'ongoing abuse' A lobbying trade body for smaller cloud providers is asking the European Commission to impose interim measures blocking Broadcom from terminating the VMware Cloud Service Provider program, calling the decision a death sentence for some tech suppliers and an illegal squeeze on customer choice.

Better than seismometers? Fiber-optic cables could be used to detect moonquakes, offering a simpler way to gather seismic data to support future missions.

Most Ubuntu desktop users will be looking at this until at least 2028 GNOME 50 is here, codenamed Tokyo after the location of the GNOME Asia Summit 2025 , and the biggest change is in fact more or less invisible, unless you look for an options button on the login screen.

Annual billed sub scrubbed after 14 days? Expect to pay 50% of yearly price Britains competition watchdog is opening an investigation into Adobes early cancellation fees on membership plans to ascertain if it breaks consumer law.

Iran-linked attackers wiped employees' devices using Intune The US government has urged companies to better secure Microsoft Intune, an endpoint management tool that was abused in last week's cyberattack against med-tech firm Stryker.

New toggle strips away browser chrome if you want Browser maker Vivaldi has opened up a new front in the browser wars by making itself disappear.

Creative pressure forces rethink as officials step back from default data use The UK government has backed off plans to allow AI companies to access copyrighted material for free for training purposes by default.

One search engine switch to rule them all in Google's response to UK competition watchdog The UK's competition watchdog has published responses to its consultation over Google's strategic market status (SMS) covering search and search advertising services - and the tech biz is offering some concessions.

It's still a job for humans, even though bots can search logs at the speed of I/O QCon London A member of Anthropic's AI reliability engineering team spoke at QCon London on why Claude excels at finding issues but still makes a poor substitute for a site reliability engineer (SRE), constantly mistaking correlation with causation.

Perks fall short as third-party AI models rack up costs with minimal notification Complaints about Microsoft's startup credits and Azure AI Foundry keep mounting, with users reporting surprise credit card charges and invoices they never saw coming.

Strategy launched after 2020 share price crash is 24% behind target Five years after launching its rescue plan to lift ERP users to the cloud and switch them to the latest software, SAP is off target by about 2 billion, The Register can reveal.

Accuracy jumps from 76% to 90% across public pilots, while users wait nearly 11 seconds for answers More powerful large language models (LLMs) are helping make the UK government's in-development chatbot more accurate but are also slowing it down, according to the Government Digital Service (GDS).

From mild vegetarianism to full-blown haterdom, there's a label for everything Opinion Are you an AI hater, an AI vegan, or a slightly more moderate AI vegetarian? Or are you on the side of the clankers? A bot-licker, a prompt- fondler, a ChatNPC?

Micron plans to cash in, after already growing revenue $10 billion in a single quarter Autonomous cars will need 300 gigabytes of DRAM or more, and humanoid robots will need similar quantities, leading memory-maker Micron Technology to predict it has a long and happy future ahead of it.

Stitch gets voice input and an infinite canvas The term vibe coding has become associated with use of AI coding assistants to create code that expresses a developers intent, even if the results are ropey and require plenty of extra work to put into production. Googles now proudly adapted the term to describe the workings of its Stitch design tool.

Baidu joins the Chinese cloud price rise party Two more Chinese cloud giants have signalled price rises for their services, again due to the impact of AI on their supply chains.

Who knew questioning authority and signaling virtue would lead to growth? Anthropic has been killing it in the business market, success that appears to be at least partially attributable to pushback against the Pentagon.

Where are you? What are you working on? Why are you doing that? Identity access and management platform Okta announced the general availability of its Okta for AI Agents, which will give customers the ability to do three things: locate agents, see what theyre doing, and shut them down if need be.

Darksword is the second iOS exploit chain in a month A new exploit kit targeting iPhone users and stealing their sensitive data is being abused by "multiple" spyware vendors and suspected nation-state goons, security researchers said on Wednesday.

Flattery and delusional talk have negative outcomes Sometimes a compliment is no help at all. Chatbot flattery, a well-known and common problem, makes things worse for humans experiencing mental health issues.

The law is the law, no matter who tells you to break it One of your studios is about to make a game that you think will be a huge hit, and you don't want to pay the contractually required bonuses. What to do? One Korean CEO turned to ChatGPT to cook up a plan to get his company out of paying up to $250 million. It went about as well as you'd expect.

Hitachi Vantara, IBM, Nutanix, and Seagate all had something to say GTC Hitachi Vantara and Nutanix announced support for Nvidias new GPUs and software at GTC 2026, much like every other storage system vendor, while IBM integrated Watsonx and other offerings more tightly with GPUzilla's offerings. Seagate demonstrated a two-tier hybrid external KV Cache composed of SSDs and disk drives, as it did last year.

PostgreSQL, MySQL, SQL Server all handled via Database Hub, vendor says Microsoft has launched a database management tool it promises will help users manage multiple databases sharing a single SQL engine.

Interlock's post-exploit toolkit exposed Ransomware criminals exploited CVE-2026-20131, a maximum-severity bug in Cisco Secure Firewall Management Center software, as a zero-day vulnerability more than a month before Cisco patched the hole, according to Amazon security boss CJ Moses.

Residents looking to ban server farms with capacity over 25 MW Ohio residents are proposing a ban on datacenters with a capacity greater than 25 MW, the latest sign of growing opposition to massive server farms across the US.

Friends and family support techs: get ready for permission changing and batch file creating Microsoft has published a handy guide for regaining access to a C: drive borked by a Samsung application, but it isn't for the faint of heart.

What exactly is AGI? Nobody knows, but Google's AI lab is asking for help trying to define it If a bot actually achieved artificial general intelligence (AGI), how would we even know? Google DeepMind boffins have come up with what they say is an empirical, scientifically grounded framework to measure progress toward AGI, and they're looking for a few good devs to actually flesh it out.

Strong forces tempting humans out of the AI loop, and reducing the experience needed to supervise and review QCon London AI is in a dangerous state where it is too useful not to use, but where by using it, developers are giving up the experience they need to review what it does, said a speaker at QCon London, a vendor-neutral developer conference underway this week.

Good luck with that The latest release of the most widely used Linux init system is here, and between dropping init script support and AI-assisted coding, we feel sure that this release will win it yet more admirers.