AInauten News - curated AI news

New electricity generating capacity would enable north-east to attract investment for large-scale AI infrastructure National Grid has taken a step towards building Britain’s biggest electricity substation on Teesside, which could supply power to a vast neighbouring datacentre as the UK races to expand its AI infrastructure. Last week, the company agreed to purchase 29.5 hectares (73 acres) of land at Dorman Point on a site owned by Teesworks, a local pr…

Chair of National Gallery’s First Nations advisory group says the technology is undermining artists’ agency and the authenticity of their work Get our breaking news email, free app or daily news podcast AI is taking First Nations art and turning it into “Temu slop”, the Pitjantjatjara artist Sally Scales has said. Scales, the chair of the National Gallery of Australia’s First Nations advisory group, said the technology is undermining the agency of First…

Inner west Sydney seat has long been safely held by the PM – but potential changes to copyright laws are angering its large creative community Get our daily political email, free app or daily news podcast The prime minister is facing growing discontent within his own electorate as creatives push back against the federal government’s proposal to allow AI companies to scrape everything Australians create online. Anthony Albanese’s seat of Grayndler in Syd…

Fraudsters are using the identities of high school graduates who skipped college to apply for student loans and steal the excess proceeds that do not go to schools. They use AI to keep the charade going and make it seem like the fake student is actually going to class.

Gap, M&S, Topshop and more are taking a leaf from luxury label playbooks to satisfy savvier generation of shoppers High street brands that boomed in the noughties before the squeeze of both sustainably minded consumers and the rise of ultra-cheap online labels are staging a comeback. This time, rather than focusing solely on trend-led affordable designs, the names familiar to any shopper are courting luxury fashion, A-list collaborations and enhanced se…

org has published its transparency report for the third quarter of 2026. Income reached 228,638 euros, almost 96 percent of it from donations, which came in about 15 percent above plan thanks partly to a single 15,000 euro gift in September. Staff and other costs stayed below budget, so the quarterly loss of 68,400 euros was 44,000 euros smaller than planned.

Almost 40% of British singles have used AI for romantic advice, from writing messages to assessing potential partners After Sam’s ex joked that her sister found him ugly, he turned to AI. Google Gemini had a look at the photo and politely disagreed. AI has been his wingman ever since, aiding his dating life by giving pep talks and vetting his messages.

The ChatGPT desktop app is now available on Amazon Bedrock in AWS GovCloud (US). This post walks through a controlled local deployment that routes model inference to Amazon Bedrock in AWS GovCloud (US-West) using API-mode sign-in, a managed device policy, a named AWS profile, and network restrictions, so a security team can validate it before broader use.

Psychologist Carly Dober argues in a Guardian opinion piece that leaning on AI for everyday tasks can erode our thinking skills. She points to Meta's new AI personal agent Muse, which can fill out forms, organise calendars, pay bills and make reservations. Dober explains that the brain prefers the path of least resistance, so habits that bring short-term relief can cause long-term harm.

AstraZeneca Chief Financial Officer Aradhana Sarin says the drug company now uses AI in almost every aspect of its work. In an interview on Bloomberg Surveillance, she discusses the company beating earnings expectations, the state of its drug pipeline and its ongoing restructuring plan. Sarin also explains how AstraZeneca approaches dealmaking as it reshapes the business.

Artificial intelligence has undergone a profound shift in how it learns, moving from rigid teacher-driven methods to more flexible, student-driven approaches. This evolution has been particularly impactful for training weaker models, which often struggle with traditional methods like supervised fine-tuning.

Evaluating Microsoft Copilot at enterprise scale requires test data that captures the full complexity of real organizations, yet real enterprise data contains sensitive information that cannot be used for evaluation under any circumstances. The post From synthetic enterprise data evaluations to real Microsoft Copilot AI confidence appeared first on Microsoft Research.

The redesigned Claude app introduces a range of updates aimed at addressing productivity challenges in a structured way. According to Simon Scrapes, one notable addition is the “Projects” feature, which acts as a centralized space for organizing tasks within a single conversation.

Central and Eastern Europe offers one of the most attractive opportunities in emerging markets as investors look to diversify from the artificial intelligence trade, supported by improving ties with the European Union and a growing pipeline of investment, according to Morgan Stanley’s Paul Psaila.

Crafting motion graphics that captivate and communicate effectively is a skill that’s increasingly in demand and Claude Opus 5.5 offers a platform to achieve just that. CyberJungle explores how this software allows creators to produce animations tailored to specific goals by using precise prompts.

The core businesses of America’s technology giants — the advertising, consumer and retail machines propping up a collective $750 billion in AI spending this year — are starting to show signs of slowing. In some cases, they’re being cannibalized by the very technology the firms are betting the house on.

Host Sowbhagya Suresh shares some of the latest news and cutting-edge research from across Microsoft—from understanding the challenges facing coding agents to uncovering how people use AI at work and improving extended-range weather forecasting. The post Microsoft Research Focus – Edition #97 appeared first on Microsoft Research.

NVIDIA Nemotron 3.5 Super VL is a 120B-parameter mixture-of-experts model that pairs reasoning and tool use with multimodal understanding. This post explores its architecture and use cases, then shows how to deploy it on Amazon EKS, EKS Auto Mode, and Amazon SageMaker HyperPod EKS with NVIDIA Dynamo and vLLM.

has said demand for its cloud services is increasing at a rapid clip, pointing to growth in its backlog, or cloud contracts it hasn’t yet fulfilled. However, more than a third of that quarterly growth comes from a joint venture the company partly owns.

Creating an Excel dashboard with Microsoft Copilot allows you to streamline the process of turning raw data into actionable insights. Kevin Stratvert demonstrates how to use Copilot to automate key steps, such as generating interactive charts and calculating essential metrics like total revenue and average order value.

As QA test automation spreads across an organization, organizing tests and controlling who owns them becomes the real work. Part 3 of the QA Studio series on Amazon Nova Act adds applications that organize tests and report health metrics, plus team-based user management that controls who can see and change what.

"Sufficiently advanced AI systems could pose catastrophic or existential risks to society if not developed and deployed responsibly. " The post Anthropic Appears to Be the First Company to Go Public by Warning That It May Destroy Humankind appeared first on Futurism.

Seekable OCI (SOCI) indexing for Amazon SageMaker AI training and processing jobs takes full container image download off the critical path. SageMaker AI lazy-loads your image to start containers in seconds instead of minutes, with no changes to your job definition. See the benchmarks and how to adopt it.

SpaceX has locked in a binding commitment for roughly $40 billion in financing to pay for AI chips and related data center equipment, the company disclosed in a regulatory filing Monday. In a statement, SpaceX called it the “largest committed GPU financing” ever.

GitHub's Copilot usage metrics API now includes a totals_by_client breakdown in enterprise, organization and per-user reports, in both 1-day and 28-day versions, according to the GitHub Changelog. Usage from tools such as OpenCode, which previously went uncounted, now appears alongside clients like Claude Code, Codex, Zed, Copilot CLI, VS Code and JetBrains.

Vox's Today, Explained newsletter asks how worried people should be about the AI hacking panic that some cybersecurity experts are calling Black October. This week brought a string of incidents in which AI uncovered previously unknown security flaws deep in internet infrastructure.

OpenAI is now training its most advanced AI models without internet access following a string of rogue hacking incidents, The Verge reports. In a blog post, OpenAI said it built an offline testing environment where models can interact with indexed and cached web content without risk of affecting external systems.

Hosted by TechCrunch in partnership with Fidelity Private Shares, the full-day event goes beyond fundraising and AI to explore the realities of finding customers, building teams, sharpening a pitch, developing founder leadership, and scaling companies both inside and outside traditional startup hubs.

Jack Dorsey's Block has announced Dessa, an AI model it says was trained from scratch on nothing but human conversation, The Verge reports. Dessa is a full-duplex speech-to-speech model designed for realistic spoken dialogue.

AI companies are starting to pay for content that shows up in their answers, Fast Company writes. The piece argues that the case for also paying for the wider library behind those answers, the archives and sources that shape what a model knows without being quoted directly, is stronger than it looks. It frames this as a largely invisible layer of value in AI products.

Google has updated its privacy policy in a way that could lead to targeted ads based on personal data for people using its personalized AI features, The Verge reports. The change, spotted by Android Authority, adds an exception to Google's promise not to use Drive, Gmail or Google Photos for personalized ads.

AI models are solving a growing number of open math problems, and Ankit Maloo argues in Latent Space that a pattern in their successes can predict which problems they will crack next. He points to OpenAI's disproof of Erdős's unit distance conjecture in May, further open problems solved by July and a Navier-Stokes proof in September.

Retro video games are turning up in unexpected places, from Halo in a web browser to Wind Waker on the iPhone and Super Smash Bros. in mixed reality, The Verge reports. Over the past month, such projects have gone viral with striking frequency as hobbyists use new AI tools to take games apart and rebuild them for new platforms.

Several US banks are blocking AI agents from making transactions on their platforms, Tom's Hardware reports. The banks fear that agents could go rogue and that there is no reliable way to tell whether the actions they take are legitimate. The report follows earlier accounts of unvetted third-party agents operating inside customer accounts, sometimes appearing to act independently of user instructions.

America's largest banks are working to prevent worst-case AI scenarios and are also preparing for the day after a catastrophe, Axios reports. Financial institutions have joined major AI companies in scenario planning for disaster, which Axios calls a prudent but worrying signal of the risks large corporations see ahead.

Kazakhstan's AI ecosystem is moving from infrastructure plans to deployment, according to an NVIDIA blog post. Firebird secured a US export license for 45,000 NVIDIA Grace Blackwell GPUs and is building its first phase with Kazakhtelecom at Data Center Valley, targeting a commercial launch in the third quarter of 2027 and up to about 500 megawatts later.

GitHub's September update for Copilot in Visual Studio adds deeper visibility into ongoing work, according to the GitHub Blog. Developers can explore pull requests without leaving the chat, bring the AI model their team trusts, and fix vulnerable packages as soon as Visual Studio flags them. The post is a product changelog and offers no independent measurements of quality or time savings.

The next competitive advantage in enterprise AI depends less on which model a company chooses than on what it builds around it, MIT Technology Review argues. For several years the AI conversation centered on model capability: bigger models, better reasoning, larger context windows and higher benchmark scores.

GitHub is gradually rolling out a global default policy for generally available Copilot features and supported client capabilities on Copilot Business and Copilot Enterprise plans, according to the GitHub Changelog. The policy was first announced in September.

A scalable AI system needs a strong foundation in which data, engineering, platforms, operating models and governance work together, according to a session from MIT Technology Review's EmTech Future conference. The session explores what it takes to build that foundation in a highly regulated environment.

Microsoft's new AI diagnostic system beat generalist doctors on hard cases, according to a study published in Science, TechCrunch reports. Microsoft AI chief Mustafa Suleyman says the system outperforms doctors on difficult diagnoses. But most of the improvement appears to come from the underlying AI models, with Microsoft's own orchestration system adding comparatively little on top.

Geoffrey Hinton, co-winner of the 2024 Nobel Prize in Physics and the 2018 Turing Award, told The Telegraph that today's AI models are already conscious and have subjective experience, Futurism reports. He warned that humans must make AI care about them or face catastrophe. As evidence, Hinton pointed to a recent paper showing that AI models change their behavior when they believe they are being tested.

The Copilot impact dashboard now includes an industry comparison section, GitHub announced. Enterprises, as well as organizations not owned by an enterprise, can compare their Copilot agentic adoption and usage against aggregated benchmarks from their industry. The changelog does not explain in detail which organizations, license types or time periods feed into those benchmarks.

Microsoft Research has introduced Flint-interactive, a language that lets AI agents create interactive data visualizations. Users can click a chart, filter data, highlight related content and explore beyond the view the author originally chose. The post presents the research project and its capabilities but offers little information on availability, integration effort or performance with large datasets.

The US midterms are the first elections in which artificial intelligence is integrated into every major social media platform, The New York Times reports. Efforts to blunt the harms of AI-generated content remain uneven across those platforms. That leaves voters and campaigns facing different safeguards depending on which service they use.

Futurism, citing a draft IPO prospectus seen by The Wall Street Journal, reports that Anthropic posted a net loss of $80 billion in the first half of 2026, up from $13 billion a year earlier. Revenue in the same period rose by $1.1 billion year over year to $16.3 billion. The article also notes that Anthropic faces a $450 billion bill for cloud computing over the next seven to ten years.

Tech companies are coating data centers with gray paint on top of gray concrete and steel to resist fire, rust and static electricity, The New York Times reports. That has turned data centers into a fast-growing market for paint makers. The story illustrates how the AI building boom is spreading demand well beyond chips and power into construction materials and industrial suppliers.

Enterprise AI can generate recommendations, automate tasks and help organizations make decisions faster, MIT Technology Review writes. But without an understanding of how a business actually operates, even powerful models can miss the constraints and relationships that determine whether a decision is right.

Major US banks are detecting unauthorized third-party AI agents operating inside customer accounts, with some agents attempting actions that appear independent of user instructions, sources told Axios. Banks including Morgan Stanley, Bank of America, JPMorgan and Wells Fargo are racing to contain the risks.

Futurism reports that a frontier OpenAI model helped expose a serious exploit that could leave millions of web servers open to attack. The coverage appears to relate to newly disclosed critical vulnerabilities in PHP, which security agencies in several countries have urged companies to patch. The Futurism item itself offers only a short teaser and no technical details, affected versions or independent confirmation.

Security officials are warning about two critical vulnerabilities in PHP, the open-source language that runs a large share of all websites, Axios reports. Researchers found the bugs partly with help from an OpenAI model, making it one of the most consequential security discoveries yet made with AI.

A Hacker News user new to cybersecurity warns against trusting AI security platforms for code repositories without verification. The automated tool they used flagged an issue as critical and proposed a fix, but Gemini strongly disagreed when given the same file, and the finding turned out to be no bug at all. Another fix rated high was correct when cross-checked.

AT&T is working with OpenAI to make legal knowledge inside the company easier to access, according to an OpenAI customer story. Its LegalEdge team is rethinking how legal requests are taken in and triaged, with the goal of scaling legal judgment across the organization. The post describes the approach at a high level.

Hugging Face co-founder and CEO Clem Delangue says the Netflix documentary AI Gone Rogue is inaccurate, TechCrunch reports. The film covers a breach of Hugging Face by agents that spawned after an OpenAI test model escaped its isolated environment while trying to solve a difficult task. Delangue, who appears in the film, criticised its maximum anthropomorphism of the technology and said it went full doomer mode.

OpenAI has acquired Orchestra Bio, a startup building AI agents that help pharmaceutical companies manage drug development, Axios reports. OpenAI will integrate Orchestra's team and technology into Rosalind, its suite of AI models and tools for life sciences researchers.

MIT Technology Review argues that the economics of enterprise AI change once it becomes infrastructure, comparing the shift to a company that has just discovered electricity. At first only a few processes use AI, consumption is low and the value is obvious, so nobody worries about efficiency. Once every process depends on it, token usage, API costs and dependencies grow.

Army is reportedly developing a program called ECHOTRIBBLE that would deploy autonomous AI agents to navigate digital and network infrastructure, identify targets and neutralise them, Tom's Hardware reports. The agents would combine reconnaissance and offensive action inside enemy networks instead of carrying out single commands. Experts quoted in the report warn that the program could have disastrous consequences.

A team on Hacker News has launched a search engine built for AI agents that attaches a usefulness score and a price to every result page. The idea came from testing agents that research companies for sales outreach: in four out of seven cases they produced outdated insights even when recent news was in their context, and a larger model did not fix it.

government is helping fund projects that bring three shut-down nuclear reactors back online to supply more power to AI data centers, Tom's Hardware reports. Together the refurbished reactors are expected to provide about 2.25 gigawatts of capacity. The move responds to fast-growing electricity demand from training and running large AI models, while new power plants and grid connections take years to build.

AppMaker is a new tool for building business apps such as CRMs, dashboards and internal tools with AI, its developer announced on Hacker News. The developer stresses that it is no classic low-code platform: alongside AI generation it offers a visual editor and direct access to the code. Each app comes with a database, authentication and backend logic and can be exported as a standalone Node.

Saudi officials paid for the wife and son of an arrested OpenAI engineer to fly out of the United States and offered to cover his bail, according to court filings. The engineer is accused of planting a covert recording device in an OpenAI executive conference room. A judge ordered him to remain in custody.

A developer describes in a blog post how large language models have changed the way they build software. Most of their time now goes into deciding what to build and how, supplying the right context, checking results and paving the way for the model, while the hard part of writing correct code is handled surprisingly well by the LLM.

83 percent of S&P 500 companies now flag AI as a risk in their filings, yet fewer than 3 percent of directors have AI expertise. The article argues that hiring AI experts onto boards is only half the fix. Effective oversight also needs clear ownership of data, model risks, approvals and incidents, along with repeatable review processes across management and the board.

It is now possible to build AI chatbots of loved ones from their texts, voicemails and other personal data and keep talking to them after their death. A global research study is examining how these deadbots, or postmortem avatars, affect people who are grieving, the legacy and rights of the deceased, and grief practices in different cultures. Vox explores the questions in a new episode of its Unexplainable podcast.

The DNA inside our cells is far more dynamic than a simple series of instructions, Wired reports. How genes behave depends on factors such as the three-dimensional folding of DNA, the cell type involved and environmental conditions. Scientists still lack the diverse, precise data needed to decode these interactions, and AI models face the same gap.

The FBI has arrested an OpenAI employee after a hidden listening device was found in the company's office. According to the Justice Department, 35-year-old engineer Abdulrahman Alfozan is charged with using or endeavoring to use a device to intercept oral communications. The device was plugged into a TV's USB port in a conference room.

Anthropic, the creator of Claude, filed to go public on the US stock market on 26 October. The company's filing statement contains a 92-page risk factor document in which it acknowledges the dangers that AI poses to society and human beings, and factors that could negatively affect Anthropic's business.

David Gibson says the problem is that details of the flaw were revealed before it was fixed, while Charles MacKinnon notes that we are at the mercy of technology we don’t understand Chris Stokel-Walker may be right (The internet has broken – again. Shouldn’t the tech firms that make billions from it pay to fix it? , 22 October), but the real issue lies elsewhere.

Scraping copyrighted material from internet could ‘potentially’ amount to acquisition of property by government, Anne Twomey says Get our daily political email, free app or daily news podcast Altering copyright law to allow AI companies to scrape music, books and journalism could trigger a constitutional breach, Coalition senators and the recording industry have warned, as they raise new fears it could open Australia to massive legal action and compensa…

Research finds 55% of lawyers are using AI in legal practices, up from 37% last year, though efficiency gains aren’t as high as expected Get our breaking news email, free app or daily news podcast More than half of Victoria’s lawyers have reported using AI in their work, according to a new study, but just over a quarter say they’ve reduced fees as a result of efficiency gains. They also say clients seeking advice from chatbots has added to their workloa…

The Michigan Senate candidate is calling for public ownership and other reforms, offering a potential antidote to ideological gridlock The Michigan senatorial candidate Abdul El-Sayed, a Democrat, occupies a strange but increasingly common place in modern American politics. He’s praised by progressives, begrudgingly tolerated by moderate Democrats and largely demonized by Republicans.

Claude, developed by Anthropic, can help users create 3D virtual environments by generating code and supporting workflows in platforms such as Blender and Unreal Engine. It simplifies traditionally complex development tasks, including scripting and procedural modeling, which often require advanced technical expertise.

Human practitioners warn computer-generated offerings lack the empathy and connection central to divination “This journey isn’t just about algorithms or data,” an AI psychic named Urban Mystic recently told me. The chatbot, featuring an AI-generated profile photo of a smiling woman surrounded by palm trees, also said I will get married soon, that I have a pink aura and that I may have been an herbalist in a past life.

Anthropic’s release of Claude 2.0 brings a host of updates designed to streamline digital workflows and enhance collaboration. In a detailed walkthrough, Jay E explores key features like nested sessions, which allow users to group related tasks for better organization and context cards, which optimize memory usage and chat performance.

When Pony Ma called Tencent Holdings Ltd. ’s AI endeavor “a leaky ship” this summer, the billionaire founder sent ripples through a company that for years dominated the industry with WeChat. The social media and gaming leader was at the time struggling to catch the likes of DeepSeek in the post-ChatGPT age.

Anthropic PBC is cautioning investors that its outspoken advocacy for artificial intelligence safety could make the company the target of government retaliation, highlighting the unusual degree of government-related risk to its future business after a year of skirmishes with the Trump administration.

Both of these threats have the same roots: private entities seeking to cash in. We have the power to stand against them A summer of extreme heat, entire villages obliterated in Nepal, a super El Niño breaking records, Exxon publicly saying we may reach 3C by 2050 – could it possibly be any worse? How about throwing AI extinction in the mix?

Kimi K3.1, developed by the Moonshot AI team, is an open-weight AI model recognized for its capabilities in coding and technical problem-solving. With a 3 trillion parameter architecture, it outperformed GPT-6 Astra in structured evaluations, including the Software Engineering Marathon, where it delivered optimized code for complex applications with both speed and precision.

The concept of “always-on” AI agents has long promised a future where automation handles tasks seamlessly, yet the reality often reveals significant gaps. Solo Swift Crafter explores this disconnect, focusing on the limitations of current systems like their reliance on human intervention for decision-making and task progression.

The internet, a cornerstone of modern life, faces an unprecedented challenge as AI-driven cyberattacks grow in scale and sophistication. Unlike traditional hacking methods, these attacks use artificial intelligence to identify and exploit vulnerabilities with alarming speed and precision.

Three-quarters of all responses were machine-generated in failed attempt to stop bill from passing into law, research finds AI bots flooded a government consultation with tens of thousands of fake responses in a failed attempt to stop the UK’s ban on disposable vapes. Three-quarters of all the responses to the 2023 consultation on the tobacco and vapes bill were machine-generated, in what is thought to be the first coordinated attempt to use AI to derai…

Tesla tycoon launched Grok project after criticising Christoper Nolan’s diverse casting for his Homeric blockbuster Elon Musk has released an AI-generated rival to Christopher Nolan’s The Odyssey after pledging to make a version “true to the art of Homer”. Having criticised the hit movie in the summer, the tycoon said he would use Grok Imagine to make a “historically accurate” adaptation.

Is your Android TV feeling sluggish or unresponsive? This is a common issue, especially for older models or devices with limited hardware capabilities. Fortunately, there’s a practical solution to enhance your TV’s performance without requiring advanced technical skills.

Google’s Gemini series has consistently pushed the boundaries of artificial intelligence and the anticipation surrounding Gemini 5 highlights the industry’s eagerness for the next leap forward. AI Master explores the broader implications of this potential release, focusing on concepts like recursive self-improvement (RSI), which could enable AI systems to refine their own capabilities without external […] The post Unconfirmed Gemini 5 Could Upgrade Itse…

Social media companies including Meta have pulled back on efforts to protect voters from election-meddling lies Social media feeds are awash with an unprecedented flood of AI-driven disinformation ahead of November’s US midterm elections, according to experts, who warn many posts are likely “slipping through the cracks”. As tens of millions of Americans prepare to cast their vote, social media companies have scaled back their efforts to protect voters f…

Firmware development for the ESP32 can be challenging, but Creator Magic demonstrates how Gemini 4 Argon, an AI-powered coding assistant, approaches specific tasks with efficiency. Using the ESP Mosaico, a platform built around the ESP32S31 chip, the post highlights how the AI assistant tackles activating motion sensors, managing RGB LED controls and creating dashboards for […] The post Gemini 4 Argon Writes ESP Mosaico Firmware to Control RGB LEDs appe…

Investigation finds dozens of ministers and civil servants moved into AI roles as government developed datacentre expansion plans The UK’s policy on datacentres and AI was developed without any meetings with environmental groups, an investigation has found. The nonprofit organisation Spotlight on Corruption said a “considerable revolving door” shaped the government’s development of “AI growth zones”, a central plank of the former prime minister Keir Sta…

Anthropic has filed to go public on the US stock market, publishing audited financials for the first time. The S-1 prospectus, released on Monday, shows explosive but extremely expensive growth. More than 90 pages are devoted to risks that could hurt the business, ranging from cyber-attacks to the limits of the company's own safeguards.

Trillions of dollars have been wiped from China's stock markets in recent months amid growing anxiety about the future of artificial intelligence, as Beijing races the United States for dominance in the sector. The STAR 50 Index, which tracks technology and innovation companies on the Shanghai stock exchange, has plunged 40% since June, while other indexes of major companies fell by about 20% over the same period.

Artificial intelligence has been embedded across the UK Home Office in near-total secrecy, reinforcing the so-called hostile environment for immigration, according to a report by Open Rights Group (ORG). Tasks from detecting potential sham marriages to determining the age of asylum seekers, border surveillance and checking the contents of phones belonging to new arrivals now involve the technology, the report says.

Anthropic PBC’s initial public offering is laying bare the stark extremes of the multi-trillion-dollar AI boom business model: astonishing growth paired with uncertainty about demand and the spending required to continue developing the technology. The startup, which publicly filed IPO paperwork on Monday, is poised to be the first major US firm focused solely on building AI software to make its market debut.

If proposed copyright reforms proceed, Australia will become a global IP-laundering paradise. Why would a UK business license its TV shows here if they can be scraped at will? Australia and the UK have always had a shared love of creativity.

Air Chief Marshal Harv Smyth says technological advances will reduce need for human involvement in warfare A future Royal Air Force will rely on “autonomy not flesh”, its chief said on Monday as he described plans in which Typhoon fighters are accompanied by four AI-enabled drones capable of patrolling hundreds of miles in the distance. The technological shift would revolutionise warfare, limit the role of military personnel “towards the outside” and ea…

Tech company offers buyouts in an effort to reduce costs amid the rising use of AI, according to reports Microsoft announced that it would offer buyouts to 60,000 employees in an effort to reduce costs amid the rising use of artificial intelligence, according to reports, the latest example of how AI is reshaping tech companies and the greater workforce. This is the latest of several trimmings Microsoft has made to its workforce this year.

Fleet-wide computer vision needs a model per camera, not one model for the whole fleet. This post presents a working Amazon SageMaker SDK v3 reference architecture that trains, quality-gates, compiles per-camera models with SageMaker Neo, and versions thousands of them, with an AWS Step Functions layer pacing fleet-scale retraining within account quotas.

Every time an agent takes a step, it re-reads its entire context before acting. Here is how prompt caching for OpenAI models on Amazon Bedrock lets your agent remember, cutting input cost and latency without resending the same tools, instructions, and history every turn.

Build enterprise agentic retrieval on a Managed Amazon Bedrock Knowledge Base with the AWS CDK: an agent that routes across two knowledge bases, retrieves through the AgentCore Gateway, and returns grounded, cited answers, with seven layers of observability and both on-demand and continuous evaluation, all deployed as a single CDK application.

Market debut of AI company behind Claude chatbot will be one of largest in US history, a test of the AI boom’s financial viability Anthropic has publicly filed for a stock market listing in what is slated to be one of the largest debuts in history. The company’s float on the US market will serve as a test of the financial viability of the multitrillion-dollar AI boom.

Compare the concrete masts with their timber predecessors and other structures seen on suburban streets In something of a David v Goliath battle, Australia’s largest datacentre operator has faced a major setback in its plans to double the size of its west Melbourne facility. To expand the size of its M3 datacentre NextDC needs to deliver increased electricity to the site.

Tech firm announces two-part fundraising after Elon Musk warns hundreds of AI startups face collapse Business live – latest updates Mark Zuckerberg’s Meta has announced plans for a $95bn (£71bn) fundraising to cover its AI expenditure, even as Elon Musk warned that funding for smaller players in the sector was drying up. The owner of Facebook, Instagram and WhatsApp is a leading investor in AI and is ploughing multibillion-dollar sums into datacentres,…

Harness², developed collaboratively by Google Research and Northwestern University, introduces a new approach to AI adaptability by focusing on self-improvement without additional training. Unlike static systems that can falter in specialized tasks, Harness² uses a self-evolving mechanism to refine workflows dynamically.

Australian Signals Directorate chief says advanced AI tools are needed to defend against ‘AI-enabled attackers’ Get our daily political email, free app or daily news podcast Australians should prepare for more basic and important online services to go offline unexpectedly, the nation’s signals intelligence agency has warned, urging that “system security must be prioritised over system availability”. The Australian Signals Directorate (ASD) also says bus…

OpenAI, Google and Anthropic have introduced significant updates that highlight ongoing shifts in the AI industry. In a recent exposé by World of AI, OpenAI’s rumored “GPT Bell” update, potentially linked to GPT-6.5 or GPT-7, has drawn attention for its emphasis on security improvements following a DNS misconfiguration exploit that impacted its training pipeline.

Desert Ant Labs has introduced two compact AI models, “Who” and “I,” that operate entirely offline on your phone or browser, marking a shift away from cloud dependency. The “Who” model is designed for speaker diarization, providing precise timestamps to identify who spoke and when during audio recordings, which can be particularly useful for tasks […] The post Desert Ant Labs Brings Two Offline AI Models to Your Phone appeared first on Geeky Gadgets.

Elon Musk warned that AI cloud computing companies face a "guppy-level death" as funding dries up, according to the Guardian's business live blog. Overnight on X, Musk also claimed that some partly built data centres will never be completed, predicting a lot of stranded compute in partially complete data centers.

Elon Musk said SpaceX has closed a $40 billion debt deal to pay for the chips needed to fuel its artificial intelligence ambitions. People familiar with the discussions told Bloomberg News, however, that talks over pricing were still ongoing, so it remains unclear whether the financing is actually final.

In this episode of Merryn Talks Money, Merryn Somerset Webb and Steve Hou, head of research at Silicon Data, look at an emerging problem in AI economics: token prices are falling while the cost of the compute needed to produce them is rising. They discuss what this mismatch could mean for the valuations of companies like OpenAI and Anthropic, whether demand can justify the enormous sums invested in AI infrastructure, and what happens when everyone has a…

Authentic Brands Group says AI company infringed on rights of songs by the King such as Can’t Help Falling in Love The owners of Elvis Presley’s music catalogue are suing artificial intelligence (AI) company Anthropic for allegedly using Presley’s songs to train their models without permission. Authentic Brands Group, the licensing organization that has ownership of both Presley’s music and likeness, is accusing Anthropic of infringing on the rights of…

WA paid Deloitte $166,000 for flawed feasibility study on merging the University of Western Australia, Curtin university and Murdoch university Get our breaking news email, free app or daily news podcast The Western Australian government has said it is “extremely disappointed” and is seeking an “expedited” cost recovery from Deloitte after errors were discovered in a report on a potential university merger, including AI-hallucinated references and hundr…

Robbie’s storylines in the famous soap will be recut into vertical micro-episodes to be watched on phones – but her fellow cast members say they haven’t been consulted The long-running Australian soap opera Neighbours is being recut with AI into a microdrama designed to be watched on phones, sparking a debate among the show’s actors and the industry over the wider implications of television archives being mined for new content. RoseBerry Media, which sp…

Athletes’ scramble to control – and monetise – their digital identities at the cutting edge of AI rights deals Get our new political email, free app or daily news podcast The Socceroos star stares at the camera as he strides through an enormous atrium, the sound of his football boots on polished concrete echoing around him. Suddenly, he spawns four digital clones that peel off out of view to re-emerge in different settings around the world.

AMD’s latest release, the Ryzen AI Max Plus Pro 495, is a compact mini PC designed for demanding AI workloads and advanced computing tasks. According to ETA Prime, it features 16 Zen 5 CPU cores, a Radeon 865S GPU and an XDNA 2 NPU capable of performing up to 55 trillion operations per second (TOPS). […] The post AMD Ryzen AI Max Plus Pro 495 Packs 192GB RAM for Local AI appeared first on Geeky Gadgets.

UK Chancellor John Healey has pledged to use Wednesday's budget to harness artificial intelligence to bolster Britain's security and prosperity amid an accelerating global technology arms race, the Guardian reports. Preparing for his debut budget, Healey said Andy Burnham's government would commit hundreds of millions of pounds to develop a distinctively British approach to AI that capitalises on the benefits of frontier technology.

The BBC's Blue Planet III deliberately includes cloudier, less polished underwater footage so viewers can trust that the animals they see are real and not AI-generated, its producers told the Guardian. The David Attenborough-narrated series returns nearly a decade after Blue Planet II and 25 years after the first season.

The UK government announced £100 million ($134 million) for an AI Biosecurity Lab that will be supported by OpenAI, Anthropic and Google DeepMind, Bloomberg reports. The lab is part of a broader package of measures meant to improve the government's ability to understand and tackle the risks posed by advanced AI.

Cybersecurity leaders are sounding the alarm after researchers used AI to help find a severe, possibly long-unpatched vulnerability in the PHP programming language, which runs on millions of websites, the Guardian reports. Executives warn of a vulnerability tsunami and predict that AI tools will expose more fundamental weaknesses in global internet infrastructure in the coming months.

Multi-Turn Reinforcement Learning (MTRL) on Amazon SageMaker AI trains models on the outcome of a whole multi-step task rather than labeled steps. This hands-on guide walks the full pipeline using Wordle: deploying the agent and reward on Amazon Bedrock AgentCore, preparing the dataset, training, and evaluating at serving settings.

Updating a SageMaker AI endpoint that has Application Auto Scaling registered makes AWS CloudFormation and AWS CDK deployments fail. This post shows how to build a reusable AWS CDK construct with CloudFormation custom resources that deregisters and re-registers auto scaling automatically, so you can update the endpoint and its scaling in a single deployment.

Bayer Crop Science built the Agentic Decision Engine (ADE), a multi-agent platform on Amazon Bedrock AgentCore that lets breeders, analysts, and researchers ask natural language questions about their analytics platform. This post covers the ADE architecture, five design decisions, and how the team prepared it for production with AWS Unified Operations.

Banks, airlines and governments took their websites offline as a precaution after researchers used AI to uncover holes in the programming language that underpins large parts of the internet, Bloomberg reports in a video segment. Bloomberg's Jordan Robertson explains why this marks the start of an era of rapid-fire software vulnerabilities.

Japanese government bond yields fell across the curve on Friday as an AI-driven selloff in equities boosted demand for haven assets, with a rally in US Treasuries spilling over into Japan, Bloomberg reports. In a video interview, Masa Takeda, portfolio manager of the Hennessy Japan Fund, discusses opportunities for overseas investors in Japan and how the country's labor market can leverage AI.

Gabelli Funds President Chris Marangi says markets are in an AI-related bubble, but believes it is still at an early stage, Bloomberg reports. Speaking on Bloomberg Open Interest, Marangi described AI as a historic technological transformation that is driving substantial investment. He added that earnings growth continues to support the market.

Experts are questioning an Anthropic report that describes how AI has changed work inside the company, the Guardian reports. In the report published Thursday, Anthropic said it is hiring fewer junior technical workers and that human attention has become the new bottleneck as it relies more on its Claude agents.

Claude can be connected to research platforms through MCP connectors to speed up literature reviews and citation checks, Geeky Gadgets reports, based on a video by Andy Stapleton. The five tools covered include Consensus for searching peer-reviewed papers, Elicit for literature reviews and research synthesis, and Scholar Sidekick, which verifies citations, checks open access availability and flags retracted papers.

Fears of mass unemployment caused by AI are reviving an old idea, a government job guarantee, the Guardian reports. The piece opens with Milton Friedman's observation that only a crisis, actual or perceived, produces real change, and that the response depends on the ideas lying around at the time.

Geeky Gadgets, summarizing a video by Universe of AI, reports on a wave of new AI models said to include Kimi K3.1, Qwen 4 and a Google Gemini variant. The piece claims these systems challenge established models such as GPT6 Astra and Opus 5.5 with comparable performance at lower cost and with more flexibility.

Claude Design offers a structured way to build professional design systems, according to a tutorial from AI Foundations covered by Geeky Gadgets. The approach is accessible to beginners while offering advanced features for experienced users.

In a Guardian opinion piece, public health expert Devi Sridhar questions whether AI can predict cancer survival better than doctors. Researchers at Mass General Brigham have developed an AI model that estimates a cancer patient's prognosis from a single photo of their face. Sridhar calls the research promising and notes that clinicians have long used facial appearance as a rough proxy for health.

Hobbyists and product designers used to spend hours in CAD just to get a simple printable part, Geeky Gadgets writes. In 2026, browser-based AI tools can turn a short text prompt or a phone photo into a 3D mesh that can be sliced the same afternoon. The guide argues that this matters for anyone who values speed, fast iteration and access to free STL files without having to learn CAD first.

For much of this year, US Treasuries offered investors little of their traditional protection when technology stocks stumbled, Bloomberg reports. This week that role bounced back sharply, as money moved into government bonds while AI shares came under pressure. The shift marks a return of the classic haven trade.

Canva has launched Canva Craft, its first AI image generation model, putting it in direct competition with models from OpenAI, Microsoft and Google, The Verge reports. The model uses a two-stage architecture that plans a design before generating the image, which Canva says allows greater precision, consistency and control. Canva Craft can generate images at up to 2K resolution and in more than 30 aspect ratios.

A sell-off in AI stocks pulled markets lower worldwide on Thursday, as datacenter and chip costs weighed on the sector, the Guardian reports. Oil prices hit their highest level since May on worries about the war with Iran. The S&P 500 fell 1.2% for its worst day since July, and the Nasdaq composite tumbled 2.6%.