Deploying AI agents in production this year? Join the early-access programme

Know what your AI agents did. Prove it.

ScrutinAIse records, attributes and reconstructs the actions of autonomous AI agents, so security, risk and audit teams can see what happened and show the evidence.

  • Transparency
  • Accountability
  • Audit
  • Investigation
Request early access See the platform

Agents now act, not just answer

Autonomous agents read data, call tools, approve transactions and change systems. When something goes wrong, the question is no longer what the model said, but what the agent did and on whose authority.

Regulators expect records

The EU AI Act sets logging and record-keeping duties for high-risk systems and their deployers; NIST's AI Risk Management Framework and ISO/IEC 42001 expect demonstrable accountability. Logs you cannot reconstruct or attest do not meet that bar.

Existing tooling stops at logs

Observability tells you an agent was active. It rarely tells you why an action happened, whether it was permitted, or lets you hand a defensible record to an auditor, a regulator or a court.

One platform for oversight of autonomous AI

ScrutinAIse sits alongside your agents and captures a complete, attributable record of what they do, then turns it into audit evidence and investigation-ready timelines.

Transparency

See every action

  • Complete action tracing across tools, data and systems
  • Decision-path capture, including inputs and context
  • Visibility of what each agent can reach and change

Accountability

Know who acted, and on what authority

  • Attribution of actions to agent, principal and mandate
  • Policy binding: what was permitted at the moment of action
  • Human checkpoints recorded, met or bypassed

Audit

Turn records into evidence

  • Tamper-evident, sealed records with integrity proofs
  • Mapping to EU AI Act, NIST AI RMF and ISO/IEC 42001 controls
  • Audit-ready reporting for internal and external review

Investigation

Reconstruct what happened

  • Forensic timeline reconstruction across agents and runs
  • Flagging of out-of-policy, anomalous or drifting behaviour
  • Root-cause analysis and incident case files

Autonomy is an accountability problem

The more an agent can do without a person in the loop, the harder it becomes to answer the questions that matter after an incident.

  • Actions span models, tools, APIs and data stores, each with its own partial log.
  • Authority is implicit: which policy, which principal and which approval applied at the moment of action is rarely captured.
  • Logs are mutable and unattested, so they carry little weight with auditors, regulators or in dispute.
  • By the time a problem is noticed, the context needed to reconstruct it has gone.

Monitoring shows you activity. Oversight proves what happened.

Talk to us about your agents

Evidence record illustrative

Agent
procurement-agent v3.2
Principal
svc-finance-ops (delegated by j.okafor)
Action
approve_po(£48,900)
Policy at time
PO-approval-v7: human sign-off above £25,000
Checkpoint
required · not met
Inputs
vendor_master#12, price_lookup(ACME-7)
Integrity
sha256 c1d0…9e2f · sealed 09:41:09Z
Every field is captured at the moment of action and sealed, so the record can be relied on later without reinterpretation.

What sets ScrutinAIse apart

Built by investigators and researchers for the moment after something goes wrong, not only for the dashboard before it does.

Evidence-grade by design

Records are attributed, integrity-sealed and preserved with chain of custody in mind, so they hold up in audit, regulatory review and dispute.

Investigation first

Timelines, case files and root-cause views are core to the product, drawing on two decades of digital investigation practice in defence, financial services and global forensic consulting.

Research-led

Founded from fundamental research in agentic AI security, digital forensics and AI governance at Keele University and the University of Liverpool, so the product keeps pace with how agents and the rules around them evolve.

Framework-aligned from the start

Controls and reports map to the EU AI Act, NIST AI RMF and ISO/IEC 42001, so evidence you collect for security is the same evidence you present for compliance.

Designed for the stack you already run

ScrutinAIse is being built to attach to agents wherever they live, without changing how they are built or where they run.

LLM API providersAgent and orchestration frameworksTool and function-calling gatewaysRetrieval and data storesCloud AI platformsSIEM and SOARGRC and audit toolingTicketing and case management

Grounded in UK research

Fundamental research at Keele University and the University of Liverpool

  • Born from active academic research in agentic and autonomous AI security, digital forensics and AI governance at two UK universities
  • Led by a lead researcher with more than twenty years in governance, risk and compliance, digital investigation and applied research
  • Practice spanning defence, financial services and global forensic consulting, with advisory roles on cyber resilience in the banking sector
  • Research-to-product pipeline: methods are published, peer-reviewed and tested before they reach the platform
  • Aligned with the EU AI Act, NIST AI RMF and ISO/IEC 42001 from the research stage onwards

Why ScrutinAIse exists

Every serious investigation comes down to one question: can you show exactly what happened, and can the record be trusted? Autonomous agents are about to make that question harder to answer, at a scale no organisation is ready for. ScrutinAIse exists so the answer is yes.

Lead researcher, ScrutinAIse

Follow ScrutinAIse on LinkedIn

Bring oversight to your AI agents

We are working with a small number of design partners who are deploying agents in regulated or high-consequence settings. If that is you, we would like to hear what you need to see, prove and investigate.

Request early access Email scrutinaise@zohomail.eu

Security researchers: report vulnerabilities to scrutinaise@zohomail.eu with the subject “Security report”.