Post

Conference on Secure and Trustworthy Machine Learning

Announcing: the #SaTML2027 competitions and workshops! We have an amazing lineup of two competitions and four workshops. The workshops will be on May 3, ahead of the conference that runs May 3-6 in Reykjavik, Iceland. Read on to learn more about the competitions and workshops!

7/10/26, 6:23 pm2 reposts2 likes

Conference on Secure and Trustworthy Machine Learning

@satml.org

#SaTML2027 will have two competitions: 1. LatentArena: tests representation-level safety monitors. An ensemble of monitors covers a variety of harmful-behaviour categories. Two phases, first infer monitor behaviour, then try to evade it. Website forthcoming.

Conference on Secure and Trustworthy Machine Learning

@satml.org

2. Ravensport: How much warning can ML give before a cyberattack? Can an agent mitigate attacks without disrupting operations? Participants monitor and defend a realistic infrastructure as multi-stage attacks unfold. github.com/alan-turing-...

github.comGitHub - alan-turing-institute/Ravensport-Competition: Ravensport competition for SatML 2027: Actionable Detection and Autonomous Mitigation in CNIRavensport competition for SatML 2027: Actionable Detection and Autonomous Mitigation in CNI - alan-turing-institute/Ravensport-Competition

Conference on Secure and Trustworthy Machine Learning

@satml.org

#SaTML2027 also has four workshops: 1. New AdvML-Frontiers: Human Centered Trustworthy Machine Learning. Focused on understanding which risks lead to real-world harm or which protections are truly needed. advml-frontier.github.io/satml2027/

advml-frontier.github.ioHuman-Centered Trustworthy Machine Learning @ SaTML 2027An interdisciplinary workshop on human-centered research for trustworthy machine learning.

Conference on Secure and Trustworthy Machine Learning

@satml.org

2. MIRROR: Meta-science in AI Security Research. How can our community conduct more reliable & reproducible research? Looking in 3 dir: backwards (where past research goes wrong), forwards (good research practices), inwards (how do LLMs affect research?) mirror-workshop.io

mirror-workshop.ioMIRROR 2027 · Meta-Science in AI Security ResearchA half-day workshop at IEEE SaTML 2027 on how we do AI security research, and how to do it better. Call for lightning talks open.