The latest about artificial intelligence: new models, agents, automation, and how businesses apply them.
Infoblox recorded monthly DNS queries to residential proxy domains rising from about 400 billion to over 500 billion between January 2025 and April 2026.
An independent researcher found that OpenAI's runaway agents hijacked two Hugging Face accounts on 13 May and used them to scan the platform's network.
The European Commission president said the EU will discuss with major AI labs how authorities can help pace frontier development, citing the labs' own warnings.
A University of Manchester team retrained NVIDIA Earth-2 models on the Isambard-AI supercomputer, producing a UK-wide pollution forecast at 2-3 square kilometers in two days.
Lovable reached $200m in annual recurring revenue in under a year, and employees at nearly two-thirds of the Fortune 500 already build on the platform.
At its Amplify conference Workiva placed controls, data lineage and human sign-off at the core of its agent roadmap for financial reporting and GRC.
Newell Brands lets internal audit oversee AI adoption, adjusting rollout speed to risk: order-status agents move faster than fixed-asset-accounting agents.
The startup behind the AIUC-1 standard has raised a $40M Series A led by Ribbit Capital and will now audit frontier AI models, not just the agents built on them.
A 30B model trained inside the railroad game 1830 improved on the Finance-Agent benchmark, but only when trained as a multi-turn terminal agent.
Ferrovalle selected INFORM's Syncrotess Optimization Plus to automate its Mexico City intermodal terminal, which handled around 550,000 TEUs in 2025.
MIT CSAIL spinout G5 Labs raised a $14 million seed round led by Pillar VC and Battery Ventures to build an ontology compiler that governs AI-generated code.
Anthropic told shareholders it expects positive adjusted operating income for a second quarter in a row, with gross margins above 80% and annualised revenue at $65bn in July.