Infosecurity
Survey Finds AI Agents Are Reshaping SOC Analyst Roles
The Infosecurity Magazine article reports that AI agents are prompting a broad redesign of security operations centres (SOCs), with 87% of security leaders in the UK, Canada, and the US indicating changes as they adopt agent-based tooling.
A Strike48 survey of 534 security and IT leaders at organisations with 200+ employees found that agents have moved from experimental to mainstream, with 74% saying they have deployed the technology in their SOC (up from 36% in May) and half (49%) using it across multiple use cases in H2 2026.
Rather than eliminating entry-level work, organisations are adapting L1/Tier-1 roles. The survey shows 49% are hiring for new skills such as AI detection engineering and automation, 42% are retraining L1 analysts to supervise agents, and 38% are shifting L1 staff into investigation roles, with 30% expanding teams with more entry-level analysts. The piece notes a lingering trust gap: only 30% of teams allow agents full autonomy, while 47% keep humans pre-approving actions and 23% operate within defined boundaries.
Tim Leehealy, co-founder and president of Strike48, emphasises that the models are mature, but governance and fixed workflows are still needed to realise reliable agent autonomy, arguing for narrow, deterministic processing steps rather than broad, unsupervised action. The article also references a September Swimlane report indicating residual concerns about junior analysts’ skill development as agents scale.