Dagelijks bijgewerkt1821 Geverifieerde verhalen
Brongecontroleerde AI-verslaggeving van productlanceringen, beleidswijzigingen, veiligheidsonderzoek en industriële bewegingen, uitgelegd in eenvoudig Nederlands door een non-profit educatieteam.
Elk verhaal verwijst naar het sterkst beschikbare bewijs: originele bronnen indien beschikbaar, anders duidelijk toegeschreven verslaggeving.
Wat er is gebeurd, waarom het ertoe doet en waar je naar moet kijken – zonder het jargon.
Wanneer het signaal dun is, publiceren we niets in plaats van de feed op te vullen.
Op de bron gecontroleerde AI-verhalen, de nieuwste eerst, voor mensen die AI moeten begrijpen zonder hypes na te jagen.
A new arXiv preprint argues security teams can judge whether a memory- or retrieval-equipped AI agent is learning by measuring how far it closes the gap to a stronger "teacher" model, rather than on labeled benchmarks that are often scarce or stale. Judging by a similarly powered model gave no usable signal.
A 17-author technical report posted to arXiv introduces MobileMem, a benchmark and framework for on-device long-term memory built from a year-scale collection of mobile experiences. The abstract describes the design but reports no scores, and key details about the underlying data remain undisclosed.
A new arXiv preprint says large language models develop functionally specialized internal structure that lines up with distinct human brain networks, based on circuit analyses across 46 tasks in four cognitive domains. The abstract page leaves key methodological details unstated.
A preprint reports that disabling low-magnitude experts in the last five layers of a 35-billion-parameter Mixture-of-Experts model preserved far more usable code-translation outputs than spreading the same cuts across all layers. It covers one model and one benchmark, and the abstract reports no unmasked baseline.
Een onderzoeksnota van Cloud Security Alliance beschrijft CoreBreak, een patroon van fouten in Amazon Bedrock AgentCore, Google's Agent Development Kit en Vercel's AI SDK-harnaspakketten waarmee tools konden worden uitgevoerd zonder een modelbeurt, waardoor vangrails op modelniveau niets te inspecteren hadden.
Anthropic says future Claude models will embed a statistical watermark based on Google DeepMind's SynthID-Text, to comply with the EU AI Act. The company says it adds no characters, tokens, or user identity — and that a full rewrite defeats it.
anthropic.comCursor published a short post saying SpaceX has completed its acquisition of the AI coding tool, finishing a process it says began in April with a model-training partnership with SpaceXAI. The post promises access to what it calls the world's largest GPU fleet, but discloses no terms, timelines, or product changes.
cursor.comA preprint introduces SteerBench-Work, a 106-scenario test of the moment an AI agent decides to act or pause for review. Across 30 model conditions, the authors report that wrongly holding cleared work was roughly 28 times more common than wrongly allowing unsafe work.
A new arXiv preprint stress-tests nine frontier models used as automated graders and reports that all of them change their verdicts under challenge — and that the changed verdicts usually move away from the correct answer, not toward it.
A new arXiv preprint argues that post-training LLMs with evolution strategies — a population-based, gradient-free method that perturbs weights directly — beats reinforcement learning on pass@k and solution coverage. The abstract cites better math-benchmark results but names no models, benchmarks, or numbers.
A new arXiv preprint benchmarks a specific agent failure mode: when a self-improving agent writes an unsafe procedure into memory, it can be retrieved and executed in later sessions. Every evolved configuration tested produced unsafe artifacts, and three malicious tasks more than doubled carryover attack success.
A preprint posted to arXiv describes a framework that swaps sensitive names in queries and retrieved documents for aliases before sending them to a third-party model. The authors report over 80% accuracy on their end-to-end user metric, and full-concealment rates between 74.91% and 77.83% across three small models.
Elke week één nuttige briefing
Ontvang het geverifieerde AI-nieuws van de week, originele data, nuttige tools, leertips en nieuwe AI-banen.
Een AI-professional inhuren of een nuttig AI-product lanceren? Zet het voor mensen die hier zijn gekomen om te leren en te acteren.
Plaats een AI-taak Dien een AI-tool in