SecurityHQ research finds AI agent accountability gaps
2 companies named across 3 categories, one of 3 articles referencing SecurityHQ. Previous coverage: AI training portals, cloud sandboxes, and security advisories - Week of September 28, 2026 (Sep 2026).
Companies mentioned
Best suited for
- Seniority
- Director
- Job function
- Cybersecurity / Information Security
- Persona
- Security Operations Leader
- Buyer role
- Decision Maker / Budget Holder
- Decision stage
- How to start
- Adoption curve
- Early Adopters
- Technology maturity
- Accelerated Attention
- Industry
- Information Technology / Cybersecurity / Application, Cloud & AI Security / AI Security & LLM/Model Security
Our classification, not the publisher's statement. Best suited for, not only for.
SecurityHQ and Datadog published research identifying a gap between how organizations define accountability for AI agents and their ability to establish what those agents recorded doing. The findings come from a survey titled “The Accountability Gap: The State of AI Agent Security and Governance 2026,” which polled 204 cybersecurity leaders at midsize and large organizations.
Agents are gaining broader access across enterprise environments. Of the respondents, 51% said AI agents already operate broadly across their organizations, and 93% expect agents to access enterprise systems more frequently than humans within five years. Meanwhile, 76% said their organization has formally defined accountability for AI-driven operational or security decisions. Even so, 45% have experienced an incident involving AI agents or automation in which accountability was unclear.
Only 49% of respondents said they could fully reconstruct every action an AI agent performed over the previous 30 days. Following a critical incident, 17% could reconstruct that activity within minutes, while 51% would need hours, 23% would need days, and 9% would need longer or were unsure.
Agent activity can also fall outside approved processes. Nearly half of respondents, 49%, believe someone inside their organization may already be running an unapproved AI agent, and 51% of those respondents said the activity spans multiple teams. The survey also found that 99% of respondents identified at least one category in which human review should remain mandatory, with financial approvals cited most often at 61%.
Feras Tappuni, CEO of SecurityHQ, said: “Most security teams can see that agents are working. Far fewer can explain what an agent did, in what sequence, and under whose authority. Those are the questions that matter when something goes wrong.”
Aaron Hambleton, Senior Vice President of Product & Services at SecurityHQ, said: “The fundamentals are identity, permissions, logging, and being able to follow an agent across systems that were not designed to be read together.” The full report is available on the SecurityHQ website.
Press release, provided by Globe Newswire on behalf of SecurityHQ. Read the original.