Full back catalogue
Every AI and data science story we've published, newest first. 33,424 articles across 669 pages. For the live feed, head to the newsroom or browse by topic hub.
Samsung SDS is raising **KRW 1.22 trillion** (about **$820 million**) through newly issued convertible bonds purchased by funds managed by **KKR**. The cash injection and strategic partnership give KKR an active…
Midi Health uses AI to train clinicians and scale care for menopausal women, building an internal chatbot to reshape workflows and operations. The company integrates conversational AI into provider training, patient…
Meta launched the sixth annual Sharing Memories Holocaust remembrance project on Instagram in partnership with the nonprofit Our 6 Million under the theme "A Vanishing Generation." The 2026 edition is digital-first and…
NVIDIA dramatically raised warranty payouts in 2025, paying **$894 million** in discrete GPU warranty claims versus **$81 million** in 2024, an approximately **1000%** increase. Warranty reserves expanded from **$416…
Anthropic has restricted initial access to its new `Mythos` model to a small set of partners and is working with defenders-first testing to identify and remediate vulnerabilities before a broader release. Canada's AI…
Jason Blum, founder of Blumhouse Productions, says his partnership with Meta shifted his view on AI in entertainment. After he was "destroyed" on social media for the deal, Blum concluded that **AI-generated content**…
**The Korea Times** says `AI` is all the rage and contrasts writers who emphasize technological details with those who enthusiastically celebrate its promise. The piece highlights a split in media coverage between…
Mathematicians are actively integrating large language models into their workflows while preserving direct mathematical intuition and rigor. Researchers combine statistical LLMs with symbolic tools such as **SAT**…
Interactive, an Australia-based systems integrator, is steering customers away from all-in public cloud toward a hybrid cloud posture to support growing `GenAI` demands while managing security, cost, and compliance. The…
EeroQ and Conductor Quantum built a functional proof of concept for an autonomous quantum computing lab by integrating **NVIDIA Ising** with real hardware. Using EeroQ's **electron-on-helium** test chip and Conductor's…
Agentic AI is moving ERP beyond passive automation to active, decision-capable systems that monitor, analyze, and act across finance, supply chain, and operations. In finance, agentic capabilities enable automated…
Vertiv has emerged as a primary beneficiary of the AI data center buildout, converting surging demand into a **$15 billion backlog** and robust 2025 results. Revenue climbed to about **$10.2 billion** in 2025, operating…
Civil resistance against widespread **generative AI** is evolving beyond boycotts and strikes to include **data poisoning**, the deliberate insertion of misleading or corrupted training data to degrade model…
**5-step** workflow walks content teams from prompt discovery to content creation, showing how to track AI visibility and optimize content using **Moz Pro**. It frames content strategy around the newest evolution of…
Anthropic launched a research-preview feature that lets developers schedule and run repeatable automations on **Claude Code** using Anthropic's hosted infrastructure. Routines can access repositories and connectors so…
OpenAI is scaling its Trusted Access for Cyber (TAC) program to thousands of verified individual defenders and hundreds of teams, ahead of rolling out more capable models. The company is introducing additional…
Suraj Singh and coauthors introduce **DIPLI**, a multi-frame, single-image-training framework that extends the Deep Image Prior paradigm to resolved astronomical targets. The method combines **Back Projection**…
**DBA** polls show **AI** adoption in the legal profession is reshaping courtroom practices and triggering concern among practitioners. The polls also place rising stipend demand at the centre of membership discussion,…
UBS analysts say the latest models from Anthropic and OpenAI are increasingly acting like application software providers and pose a material threat to incumbent enterprise software vendors. After attending the HumanX AI…
PostgreSQL 18 and accompanying field guidance argue that most AI failures are not model problems but data and integration seams. Keeping embeddings and relational truth together-using `pgvector` inside **PostgreSQL 18**…
Transfer learning, de-novo machine learning, and ensemble integration improve five-year survival prognostication for early breast cancer relative to the pre-trained `PREDICT v3` tool. Models were trained on the…
AWS prescribes secure access patterns for AI agents and coding assistants using the **Model Context Protocol** (MCP). The guidance treats MCP-connected agents as dynamic, potentially high-privilege actors and recommends…
Anthropic changed **Claude Enterprise** billing from a fixed per-seat subscription to a lower headline seat fee plus mandatory consumption commitments. The new commercial design surfaces a **$20/month** seat for…
South Korea recorded the highest number of AI patent filings per capita in 2024, with **14.31 AI patent filings per 100,000 people**, according to the **2026 AI Index Report** from Stanford HAI. The country also posted…
The financing for a massive AI data center campus in Saline Township, Michigan, built for **OpenAI** with **Oracle** as anchor tenant, has ballooned to **$16 billion**. Reports show **PIMCO** is negotiating to supply…
Enterprises are shifting to a customer journey-led enterprise architecture to make AI-driven customer experience (CX) work. Fragmented systems and siloed data create what the report calls **service debt**, which drives…
Crypto exchanges Coinbase and Binance are pursuing access to Anthropic's new security-focused model, **Mythos**, as custodians brace for a new era of AI-enabled vulnerability discovery and exploitation. Anthropic is…
Baidu released **ERNIE-Image**, an open-source **8B** diffusion transformer that targets cases that break other generators: legible in-image text, strict layouts, multi-panel comics and bilingual prompts. The release…
SK Telecom's early **$100 million** investment in **Anthropic** in **2023** has appreciated sharply as the AI startup scales revenue with `Claude` and `Claude Mitos`. SK Telecom holds about **0.3%** of Anthropic,…
Crypto exchanges and custodians, including Coinbase, Binance, and Fireblocks, are pursuing limited access to Anthropic's new model, `Claude Mythos`. Anthropic has restricted `Claude Mythos` to vetted partners because…
Jamie Dimon, CEO of JPMorgan Chase, said Anthropic's new model Mythos has surfaced more software vulnerabilities after JPMorgan tested the `Mythos preview`. Dimon called AI a double-edged sword, saying "AI's made it…
Procter & Gamble is applying advanced computer vision, synthetic data generation, and 3D human modeling to understand human behavior at product scale. Oya Aran, Director of R&D Data Science and AI at P&G, will present…
The **UK AI Safety Institute** published its independent evaluation of **`Claude Mythos Preview`**'s cyber capabilities, conducting an original analysis of the model's behavior in cyber contexts. The report frames…
AI-driven dynamic pricing is reshaping how clothing is priced online, with retailers using real-time signals to raise or lower prices based on demand, inventory, competitor moves, social trends, and individual shopper…
A correction published in **JMIR Med Inform** clarifies misprinted y-axis labels in figures from the 2023 paper "An End-to-End Natural Language Processing Application for Prediction of Medical Case Coding Complexity" by…
**Cochrane** plain language summaries (PLSs) aim to make systematic review findings accessible to the general public, but inconsistencies exist. This retrospective observational study classifies Cochrane PLSs by…
Om Malik argues that society and the tech industry treat human mistakes as inevitable but punish machine errors harshly. Recent incidents, including an accidental public release of internal files by **Anthropic**, a…
Two engineers built **The Gas Index**, an AI-driven national gas-price tracker that fills gaps left by Google Maps by combining public data, crowdsourced photos, and automated calls. The system uses image OCR to read…
GitHub launched Season 4 of the **Secure Code Game**, a free, open-source, in-editor training that teaches developers to find, exploit, and fix vulnerabilities in **agentic AI**. The season focuses on securing `Agentic…
Anthropic won a high-profile legal and reputational victory by refusing Pentagon demands to permit autonomous lethal systems and domestic mass surveillance, but that principled stance has a business cost. Viral consumer…
CoreWeave posted **168%** revenue growth in 2025 and holds a **$66.8 billion** backlog, driven by long-term, take-or-pay style contracts and enterprise demand for GPU capacity. NVIDIA has committed **$2 billion** and…
NVIDIA's stock rallied more than **18%** over a recent ten-day stretch, its longest winning run since 2023, after CEO **Jensen Huang** said the company has **$1 trillion** of GPU orders through 2027. The order pipeline,…
**Microsoft** agrees to rent **30,000** **Nvidia** `Vera Rubin` chips from **Nscale** at a site in Norway that was initially intended for OpenAI and marketed as part of **Stargate**. The agreement directs large-scale AI…
On **April 14, 2026**, Oracle Financial Services launched an agentic AI platform extension that embeds pre-built AI agents into corporate banking workflows. The suite includes `Loan Data Extraction Agent` and `Loan Data…
nCino has rolled out the Analyst Digital Partner, a role-based, agentic AI that customers report reduces commercial relationship review time by **60-70%** and can be deployed in as little as **36 minutes**. Built on…
Amazon SageMaker HyperPod delivers a managed, Kubernetes-orchestrated platform for large-scale generative AI inference. HyperPod integrates with Amazon EKS, SageMaker JumpStart, S3, and FSx for Lustre to enable…
The article explains what social media monitoring is and how it works, describing methods to track brand mentions, measure sentiment, and analyze engagement signals. It outlines practical steps to turn those insights…
A domain-optimized large language model, DeepSeek-R1, was evaluated on Chinese emergency radiology reports to detect reporting errors under time pressure. Researchers assembled **7,435** reports spanning radiography,…
Sebastian Mallaby argues the United States should pivot from containment to cooperation with China on AI safety. He reports Chinese firms have sidestepped US chip export controls by renting offshore data-center…
The 2026 Executive Brief from Phocuswright and ITB Berlin projects that by 2046 artificial intelligence, concentrated data control, and shifting trust mechanisms will fundamentally change global travel. The report,…
