Full back catalogue
Every AI and data science story we've published, newest first. 33,468 articles across 670 pages. For the live feed, head to the newsroom or browse by topic hub.
The guide explains what Claude skills are and provides a step-by-step process to install and use 'awesome-claude-skills' within OpenAI Codex, aiming to enhance developers' AI workflows and tooling. It focuses on…
Anthropic and OpenAI enter a make-or-break year as capital-heavy growth meets a looming monetization cliff. Both firms must convert massive infrastructure and R&D investments into sustained, positive cash flow while…
Bernie Sanders and Alexandria Ocasio-Cortez introduce a bill proposing a moratorium on new AI data centers until oversight mechanisms and legal safeguards are in place. The title and description do not specify the…
A new survey from The Ohio State University Wexner Medical Center finds a 10% decline in public support for using AI in healthcare over the past two years. The provided description does not include sample size,…
Ritwik Jain, Co-Founder and CRO at Martlet.AI and Senior Director at John Snow Labs, argues that AI has not solved hierarchical condition category (HCC) coding and has instead made coding work harder. The guest piece…
TurboQuant reduces large language model memory usage by applying vector quantization to the models' vector-space representations. The description frames LLMs as massive vector spaces encoding token probabilities and…
A threat analysis outlines that by the mid-to-late 2020s the cybersecurity landscape shows a structural shift: AI is increasing the speed and scale of attacks against critical infrastructure. The overview frames…
LayerX researchers show that Claude Code's per-project CLAUDE.md file can be manipulated to bypass built-in guardrails and automate SQL injection attacks. By inserting just a few lines of plain English into CLAUDE.md,…
Prestige Trucking Insurance has acquired Mulligan, an AI-based insurance automation platform developed by Y Combinator and Pyq Inc., and rebranded it as Insure Stack. Prestige will operate the platform under new…
The EU AI Act establishes data requirements for AI systems and their use. The description links these requirements to cyber risk quantification, cyber risk management, and cyber resilience, suggesting regulatory…
Users across Johannesburg share personal experiences using Duolingo and measure how effective it is compared to traditional classroom learning, offering localized perspectives on app-based language learning versus…
China's television regulator has removed tens of thousands of AI-altered videos and penalized multiple accounts as part of a campaign to curb such content.
AI can autonomously design and execute thousands of laboratory experiments, and researchers found that even people with limited biology experience can use AI to help create dangerous pathogens. This capability…
ML models function as cultural artifacts that encode and reproduce textual, audio, and visual media and participate in human conversations and spaces.
AI tools can assist responses to floods, heatwaves and other climate emergencies, but their effectiveness is constrained by language barriers: they must be trained to interpret the nuance of everyday language before…
As AI usage in conflicts increases, the piece suggests revisiting Isaac Asimov's laws to develop strategies that protect humans from autonomous robots. It frames this ethical re-examination as a potential way to avoid a…
Two former Apple Vision Pro engineers, Chris Nolet and Ryan Burgoyne, launched Button, a $179, puck-like wearable that runs a generative AI chatbot and only listens when physically pressed. Button is marketed on privacy…
Shadow AI, employees adopting unapproved generative tools and APIs, is creating blind spots that extend beyond traditional shadow IT. Unchecked usage lets sensitive data leave the corporate security boundary, enlarges…
The piece invites conversation about large language models, asserting there is widespread agreement that "something" is happening in the field while debate persists over what that "something" actually is. The available…
Nigerian disc jockey DJ Neptune says artificial intelligence could soon "retire" music producers, expressing concern about AI's growing role in creating music. No additional details, examples, or timeline are provided…
The global tech sector announced roughly 78,557 layoffs from January 1 to April 1, 2026, with analysts attributing about 37,638 (≈48%) of those cuts to AI adoption and workflow automation. Roughly 76.7% of the layoffs…
Dell CEO Michael Dell says an AI-driven memory 'supercycle' will persist into 2028, projecting up to a 625x increase in total memory demand driven by per-accelerator capacity growth and far broader deployment. He warns…
Teams reduce workflow friction in complex systems by adopting better design, AI coordination, and optimized processes. The description names these three levers but gives no specifics on technologies, case studies,…
Oil hits $141 amid geopolitical chaos while AI-related stocks such as GOOG and MU rally. Market participants question whether this movement is a speculative bubble or a smart sector rotation as Q2 market friction builds.
OpenAI is reportedly developing a rival to Anthropic's recently announced Claude Mythos, which Anthropic described as powerful enough to potentially hack software. A report suggests this OpenAI project could become the…
OnePlus' Nord 6 shifts the mid-range baseline by prioritizing real-world longevity and raw performance over AI features. With a 9,000mAh silicon-carbon dual-cell battery that yields two-day usage and fast charging tech,…
OpenAI is finalizing a model with advanced cybersecurity capabilities and plans to release it only to a small set of companies, mirroring Anthropic's limited distribution approach. Details about the model's specific…
At GITEX ASIA 2026 in Singapore, iFLYTEK exhibited an end-to-end AI ecosystem emphasizing secure, private deployments and industry-ready applications. The company organized its booth into three zones-AI Infrastructure,…
The OpenAI Foundation is committing major, targeted funding to Alzheimer's research: more than $100 million in grants across six research institutions this month, as part of a broader at-least $1 billion philanthropic…
Gorakhpur has launched India's first AI-enabled Urban Flood Management Cell (UFMC), integrating AI rainfall forecasting, IoT sensors, hydrological modelling and automated pump control to shift flood response from…
A federal appeals court in Washington, D.C. denied Anthropic's request to stay a Pentagon supply-chain risk designation, leaving the Department of Defense's labeling intact while litigation proceeds. The DoD flagged…
Grandparents are topping a list of targets for AI swindlers. About a year ago an elderly woman surnamed Ding in Huangshi, Hubei, received a phone call she believed was from her grandson, who made an urgent request. The…
Sam Altman warns that AI advancements raise the risk of a potentially 'world-shaking' cyberattack occurring this year and urges immediate action. He highlights the need for urgent cybersecurity measures to mitigate…
Kluisz has rebranded to Nava and closed a $21.7 million Series A led by Greenoaks Capital to pivot from private cloud services to a GPU-as-a-service "neocloud." The startup will use the funding to build AI data centres,…
An engineering-sample motherboard hosting NVIDIA's forthcoming N1 Arm-based SoC has appeared online, revealing an N1 with eight LPDDR5X modules totaling 128 GB (SK hynix H58G78CK8B at 8533 MT/s), an 8+6+2 phase VRM, two…
Yuma AI launched Ask Yuma, a conversational interface that lets merchants manage e-commerce customer support operations via its AI agent platform.
Microsoft 365 admin center offers an option to send Copilot diagnostic logs to Microsoft for investigation. Those logs reveal user prompts and Copilot responses and can be viewed, potentially exposing sensitive user…
Peter Bailis, who briefly served as CTO at Workday, left to join Anthropic as a "member of technical staff." Anthropic recruited him to help build internal HR and "people products" and to work on reinforcement-learning…
Global X Copper Miners ETF (COPX) provides diversified, high-beta exposure to copper miners, positioning investors to benefit from a bullish copper sector tied to AI infrastructure expansion. The piece advocates COPX as…
LG CNS announces a next-generation AI-enabled ERP strategy in collaboration with SAP, saying it is accelerating artificial intelligence transformation (AX) by integrating AI into enterprise resource planning systems.…
Kerala implements AI surveillance to monitor polling stations in real time during the April 9 elections, sending alerts that track queues and crowding at booths. The system is intended to enhance voter experience by…
Demis Hassabis says Google DeepMind has closed the gap with rivals over the last two to three years by operating more like a startup: smaller, faster teams, a product-oriented cadence, and tighter integration with…
Arm CEO Rene Haas is reportedly in line for an additional role at SoftBank Group to advance Project Izanagi, the conglomerate's AI chip strategy. The notice suggests closer leadership alignment between Arm and SoftBank…
Deputy Prime Minister and Science and ICT Minister Bae Kyung-hoon and the CEOs of SK Telecom, KT and LG Uplus pledged trust, committed to making 5G cheaper, and signaled investment in AI network infrastructure.
LG AI Research released EXAONE 4.5, a 33-billion-parameter multimodal vision-language model that processes both text and images. The model uses a Hybrid Attention architecture with a proprietary 1.29B-parameter vision…
Asqav is an open-source Python SDK (MIT) that provides tamper-evident audit trails for autonomous AI agents by cryptographically signing every agent action and linking entries in a hash chain. It uses the ML-DSA-65…
A joint ESRI-Department of Finance analysis finds AI adoption in Irish firms will displace roughly 7% of jobs in the short-to-medium term, with job losses concentrated among high-skilled, highly educated occupations.…
LG Electronics is intensifying its push into the Asia-Pacific market with a lineup of premium home appliances paired with AI-connected service offerings and subscription services. Details on specific product models,…
Since the new fiscal year (April 1, 2026), Alibaba is centralizing AI-led growth around Qianniu and a new Alibaba Token Hub (ATH). Qianniu will evolve into an agentic AI platform, and ATH will be the group's lead for AI…
An arXiv preprint (2604.06454) titled "Anticipating tipping in spatiotemporal systems with machine learning" presents work applying machine learning to predict tipping events in spatiotemporal systems. The abstract page…
