# Detecting AI Training in Datacentre Power

DevFeed: [Detecting AI Training in Datacentre Power](<https://devfeed.tech/articles/detecting-ai-training-in-datacentre-power-96565.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2610.30182>)

Author: Tom Kimpson, Mauricio Baker, Emlyn Graham

Published: 2026-10-23T04:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>), [arXiv: Cryptography and Security](<https://devfeed.tech/sources/arxiv-cryptography-and-security.md>), [arXiv: Machine Learning](<https://devfeed.tech/sources/arxiv-machine-learning.md>)

Topics: [AI Infrastructure](<https://devfeed.tech/topics/ai-infrastructure.md>), [Advanced AI workflows](<https://devfeed.tech/topics/advanced-ai-workflows.md>), [sequential-testing](<https://devfeed.tech/topics/sequential-testing.md>)

Tags: [2](<https://devfeed.tech/tags/2.md>), [abstract](<https://devfeed.tech/tags/abstract.md>), [adversarial](<https://devfeed.tech/tags/adversarial.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-governance](<https://devfeed.tech/tags/ai-governance.md>), [ai-training](<https://devfeed.tech/tags/ai-training.md>), [alarm](<https://devfeed.tech/tags/alarm.md>), [arxiv](<https://devfeed.tech/tags/arxiv.md>), [but](<https://devfeed.tech/tags/but.md>), [cadence](<https://devfeed.tech/tags/cadence.md>), [cluster](<https://devfeed.tech/tags/cluster.md>), [compute](<https://devfeed.tech/tags/compute.md>), [computing](<https://devfeed.tech/tags/computing.md>), [costs](<https://devfeed.tech/tags/costs.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-cr](<https://devfeed.tech/tags/cs-cr.md>), [cs-cy](<https://devfeed.tech/tags/cs-cy.md>), [cs-lg](<https://devfeed.tech/tags/cs-lg.md>), [develop](<https://devfeed.tech/tags/develop.md>), [faster](<https://devfeed.tech/tags/faster.md>), [form](<https://devfeed.tech/tags/form.md>), [governance](<https://devfeed.tech/tags/governance.md>), [power](<https://devfeed.tech/tags/power.md>), [training](<https://devfeed.tech/tags/training.md>)

## AI overview

The method detects synchronous model training from facility-level power by tracking the wandering cadence of training iterations with a Viterbi algorithm. In simulations, with a meter sampling at 2 Hz or faster and against a declared inference workload, it detected at least 96% of runs at a 1% false-alarm rate when operators did not try to hide training. Of eight hiding strategies considered, only extreme variation in work per iteration evaded the tracker. Its strongest version hid training in up to 73% of five-minute observations, yet at least 87% of runs were caught over a day at a 0.1% false-alarm rate per run. It preserved throughput but, in small-scale training, required 93-267% more compute to reach the same loss. These are simulation results, and the abstract presents passive power metering as a promising component of compute verification.

## Source excerpt

arXiv:2610.30182v1 Announce Type: new Abstract: AI governance agreements require methods to detect large model training runs on a monitored computing cluster, without trusting the operator. We develop a method to detect synchronous model training in facility-level power by Viterbi tracking of the wandering iteration cadence. On simulated datacentre power, when the operator makes no attempt to hide, the Viterbi tracker detects at least $96\%$ of training runs against a declared inference workload, at a $1\%$ false-alarm rate and within $0.04$ of the Neyman--Pearson optimum, given a meter that samples at $2$ Hz or faster. Of eight adversarial strategies an operator might use to hide, only extreme variation of the work per iteration evades the tracker. Its strongest form hides training in up to $73\%$ of five-minute observations, but over a day at least $87\%$ of such runs are still caught, at a $0.1\%$ false-alarm rate per run. This adversarial strategy costs no throughput, but in small-scale training the operator needs $93$--$267\%$ more compute to reach the same loss. A modest passive meter is sufficient to make hidden training both detectable and expensive, which makes power metering a promising part of compute verification that does not depend on trusting the operator.