These last few months have seen a wave of AI security breaches, with no frontier model provider immune to the risk.Perhaps the most well-known of these is the Hugging Face incident, when OpenAI agents broke into the...
In the past six months, inference costs for frontier-level AI capability dropped roughly 85%. Open-weight models from Meta, Alibaba, and others now match last quarter’s best closed models on the benchmarks that matter....
Insights from a session I hosted at the 2026 AI Agents Conference in NYC.There’s a moment every team building agents eventually hits. Something breaks within the agent output. That might be a wrong recommendation, a...
In the early days of cloud, companies raced to migrate workloads before they’d figured out how to monitor them. Move fast and instrument later was the operating model at the time. Unfortunately, these organizations...
71% say they’re actively preparing. Only 29% have a documented definition of what that means.Earlier this year, Monte Carlo partnered with CDO Magazine to survey senior data and AI executives, primarily at large...
Moving from “what should I recall?” to “what should I notice?” and other challenges in building a proactive multi-agent systemMemories in the context of a lineage graph. Image provided by the authors.Co-authored with...