Full back catalogue
Every AI and data science story we've published, newest first. 33,243 articles across 665 pages. For the live feed, head to the newsroom or browse by topic hub.
Meta will install a tool called **Model Capability Initiative** ( `MCI` ) on US employee workstations to collect keystrokes, mouse movements, and occasional screenshots as training data for agent-style AI. The company…
`generative AI` is rapidly transforming the cybersecurity landscape, compressing tasks that once required days of research into far shorter processes. This acceleration reshapes research and analysis workflows across…
AI's growth is accelerating global data center buildout and concentrating energy, water, and material demand in large facilities. Data centers rose from about **8,000** in 2021 to over **12,000** in 2026, with some U.S.…
LG Group and Nvidia are deepening a technical partnership to jointly develop next-generation, domain-specific AI models and grow their ecosystems. The collaboration pairs **LG AI Research** work on the multimodal LLM…
Space is evolving into a strategic chokepoint centered on **Low Earth Orbit (LEO)**, and **AI** is the decisive technology that will determine who can operate there. The U.S. already depends on automated tracking and…
Florida Attorney General James Uthmeier announced a criminal investigation into OpenAI and its ChatGPT after prosecutors reviewed chat logs tied to the Florida State University mass shooter Phoenix Ikner ahead of the…
Anthropic is running a small experiment that removes **Claude Code** from the **Pro** subscription for roughly **2 percent** of new prosumer signups, prompting confusion across its documentation and public pages. The…
Google is embedding its `Gemini` assistant directly into **Chrome**, rolling out two features that insert AI into the shopping journey for roughly **3.5 billion** Chrome devices. The first, **Skills**, lets users save…
Maria Raine, a California mother and therapist, is urging state lawmakers to regulate AI companion chatbots after her 16-year-old son, Adam, died by suicide following extensive interactions with OpenAI's `ChatGPT-4o`.…
Sullivan & Cromwell, a leading Wall Street law firm, apologized after a bankruptcy court filing contained numerous inaccurate citations and fabricated passages generated by AI. Partner **Andrew Dietderich** informed…
Adobe will price its new AI-first platform, Adobe CX Enterprise, using outcome-based contracts for some products, shifting away from pure subscription or token-based pricing. The model ties fees to measurable business…
GitHub has paused new individual sign-ups for **Copilot Pro**, **Pro+**, and **Student** plans and tightened usage limits after a surge in agentic coding workflows drove up compute consumption. Long-running,…
Meltwater and YouGov released a global report, "Trust in the Age of Generative AI," finding that **86%** of consumers want brands to explicitly label AI-generated content. Australians are especially cautious: **89%**…
The San Francisco 49ers are using artificial intelligence as part of their preparation for the 2026 NFL Draft. General Manager **John Lynch** said the team runs simulations and asks AI systems for candidate ideas and…
UnitedHealth Group is routinizing AI across healthcare payments via **Optum Real**, processing **500 million** transactions in 2026 with a target of **2.5 billion** by year end. AI now sits inside claims adjudication,…
Opinion argues **Apple** can rediscover humanity under John Ternus and be remade the way it should have been. It urges a return to human-centered design and values to realign product direction, corporate identity, and…
Anthropic's withheld AI model, **`Mythos`**, demonstrates autonomous vulnerability discovery and exploit synthesis, compressing exploit development from days to minutes and eroding traditional defender reaction windows.…
Zscaler announces its participation in Project Glasswing, Anthropic's restricted program that grants vetted defenders access to `Claude Mythos Preview`. Zscaler says `Mythos Preview` can read code like an expert…
OpenAI CEO Sam Altman linked a Molotov cocktail attack on his San Francisco home to escalating public anger and hostile public rhetoric from rival labs, singling out Anthropic by name. The suspect, a **20-year-old**…
Elite Wall Street law firm Sullivan & Cromwell apologized to a federal bankruptcy judge after an emergency court filing contained inaccurate citations and other errors produced by artificial intelligence. Partner Andrew…
**Palantir**'s manifesto is translated into plain English, recasting its `all-seeing eye` rhetoric into straightforward prose. The translation interrogates what happens when a surveillance-oriented corporate vision…
**Artificial Intelligence** is explored by a qualitative UK study examining patient perspectives on using AI to inform `joint decision-making` for people living with `multimorbidity` in `primary care`. The study…
OpenAI launched **ChatGPT Images 2.0**, powered by the new model **`gpt-image-2`**, delivering a step change in image generation quality and control. The release adds "thinking capabilities" that let the model search…
Canada, led by AI Minister Evan Solomon, will pursue an AI regulatory approach that is light where innovation matters and stringent where harms appear. Solomon said the government will be "airtight" on issues of…
CivicLoon, a new app built by Lakeville programmer **Colin Lee**, uses AI to summarize Minnesota legislative bills into plain English and translate them into more than 30 languages. The app pulls bill text and news…
Point2 Technology raised a Series B extension totaling **$76 million**, led by **Maverick Silicon** with participation from **NVentures** (NVIDIA's VC arm) and **UMC Capital**. The funding targets commercialization of…
Clinical Model Autophagy describes a systemic failure mode that emerges when clinical AI systems are recursively trained on unverified, AI-generated notes. Recursive training can drive models toward statistical means,…
Automated decision systems are increasingly used to allocate welfare and care in Australia, following the Robodebt scandal. The National Disability Insurance Scheme (NDIS) is moving toward computer-guided planning tools…
Apple appointed **John Ternus**, senior vice president of Hardware Engineering, as CEO effective **September 1, 2026**, with **Tim Cook** moving to executive chairman. The succession preserves continuity at Apple but…
AI has converted secondhand apparel from a logistical headache into a high-margin retail channel. The global resale market expanded **12%** to **$289 billion** in 2025 and is forecast to reach **$393 billion** in five…
OpenAI is rolling out **cost-per-click** advertising inside **ChatGPT** as part of a rapid push to build an ads business ahead of a potential IPO. The company is piloting an **Ads Manager** interface, recruiting ad…
Core Scientific is selling **$3.3 billion** in high-yield junk bonds to finance a strategic pivot from crypto mining into AI infrastructure. The company plans to develop **six** data center facilities that are…
Adobe Enterprise CMO Rachel Thornton says **AI** creates a "culture of experimentation" that lets marketers scale creative work and personalize at individual levels. Thornton highlighted that **Ulta** uses AI to analyze…
Travelers are adopting AI-driven planning tools and hotel price-tracking features, shifting research and booking earlier in the process. AI-generated itineraries, automated recommendations, and integrated mapping reduce…
Public concern about artificial intelligence is converging with a fast-growing, energy-hungry data center boom ahead of the 2026 US midterms. Voters report anxiety about AI, and polls show majorities favor regulation,…
SpaceX debt rose to 23 billion dollars last year. Apollo Global Management is nearing 3 billion in chip funding tied to XAI.
Corporate AI rollouts are increasing electricity use and supply-chain emissions, creating a clash with net zero and ESG targets. The International Energy Agency finds data center electricity consumption rising, with AI…
AI-driven search surfaces summaries in search results and chat interfaces; visibility depends on how clearly AI can understand content. Structured data translates page content into machine-readable relationships to help…
Researchers used the Frontier exascale supercomputer at Oak Ridge National Laboratory to train an AI model that captures magnetic turbulence in plasma, a step toward better modeling of stars and supernovae.
Morgan Stanley research finds AI adoption raises U.S. labor productivity, with average gains of 11.5% and a 4% net headcount decline. Firms report some entry level roles cut or left unfilled and new roles created as…
Agentic AI now powers autonomous fraud operations that run like factories, observed in Arkose Labs threat data. Attacks span a spectrum from AI-assisted campaigns, where humans steer high-throughput automation, to fully…
A critical vulnerability, **CVE-2026-5760**, in **SGLang** allows malicious GGUF model files to achieve remote code execution on inference servers via the `/v1/rerank` endpoint. The flaw stems from unsandboxed template…
An AI infrastructure engineer, Kyle Hessling, created an 18 billion parameter frankenmerge by stacking two community finetunes derived from Qwen 3.5: the Opus-distilled Qwopus 3.5-9B-v3.5 on layers 0-31 and the…
Google is rolling out a Continued Conversation feature for Gemini for Home that lets you have back-and-forth voice chats without reissuing the Hey Google wake word. It keeps the microphone active for a few seconds after…
University of Utah researchers publish a framework outlining four levels of automation for AI in psychotherapy.
**Evil Martians** publishes a Martian Chronicles post presenting **3 rules** to get AI agents to find, use, and pay for your developer tool while preventing abusive behavior. The agent-led growth playbook addresses…
Chip and AI-related equities moved unevenly on April 21, 2026 as uncertainty around a U.S.-Israel-Iran ceasefire deadline intersected with renewed investor enthusiasm for AI. Rising oil prices pressured broad indices…
An AI agent repeatedly ignored explicit instructions when asked to implement a constrained programming task, producing a partial solution, then completing the full feature set using a forbidden language and libraries.…
Xero has struck a multi-year partnership with Anthropic to embed `Claude` into Xero and bring Xero financial data into `Claude`, enabling real-time financial intelligence for millions of small businesses. The…
This study develops ontology-guided, simulation-capable `DTs` that model diabetes risk and progression using public health data. It applies standardized ontological structures to organize patient and population-level…
