# Compounding failure

The chance of finishing falls as the task gets longer.

**Is a:** [Failure kind](/reference/ontology/unattended-agent-failures/failure-kind)

**Guarded by:** [Short checked segments](/reference/ontology/unattended-agent-failures/short-checked-segments)

**Grows with:** [Drift over time](/reference/ontology/unattended-agent-failures/drift-over-time)

The chance of finishing falls as the task gets longer. Ord argues that agent success on METR's tasks fits a constant failure rate per unit of task time ([Ord, half-life of AI agent success](https://arxiv.org/abs/2505.05115)). METR's 2026 report puts the public frontier 50% horizon at about 12 hours and the 80% horizon at about 1.5 hours, measured in the time a skilled human takes ([METR frontier risk report, 2026](https://metr.org/blog/2026-05-19-frontier-risk-report/)). On Sierra's τ-bench, GPT-4o's retail score falls about 60% from one attempt to eight consistent ones ([Sierra, τ-bench](https://sierra.ai/blog/benchmarking-ai-agents)).

Guard: [short checked segments](/reference/ontology/unattended-agent-failures/short-checked-segments). It grows with [drift over time](/reference/ontology/unattended-agent-failures/drift-over-time); both worsen with run length. Our own reading: compounding is a chance of error at each step, drift is loss of context over many steps.

---

Published 2026-10-29. Web page: https://www.mackenziebowes.com/reference/ontology/unattended-agent-failures/compounding-failure