Softlander Book a call
Agent security and cybersecurity

Agentic AI threat modelling

Prioritise credible attack paths against the agent's business authority

A workflow connects untrusted information to tools that can change business state. The decision is how to prioritise credible attack paths against the agent's business authority. The work begins with your actual workflow and looks for the evidence that separates intended behaviour from a consequential exception. It is particularly relevant to security architects and agent product owners. A useful outcome is an explicit account of what was examined, what remains uncertain, and which decision the evidence can support.

Book a 30 minute readiness callBrowse this capability group

The agent security relationship

  1. Starting evidenceData-flow diagrams

    Establish the basis for controlled input.

  2. Decision boundaryTrust crossing

    Locate where external content gains influence.

  3. Decision artefactAgent threat model

    Retain the scoped finding and its unresolved dependencies.

Proposed working model. A bounded challenge cannot establish exhaustive protection.

Decision

Frame the decision around controlled input

Questions and the evidence they need

Criterion / evidenceControlled input
Inspect data-flow diagrams for the relevant constraint or observation.
Criterion / evidenceTrust crossing
Inspect actor assumptions for the relevant constraint or observation.
Criterion / evidenceConsequential action
Inspect assets for the relevant constraint or observation.
Read each criterion against the available evidence before deciding that additional work is needed.

The decision is how to prioritise credible attack paths against the agent’s business authority. This is relevant to security architects and agent product owners when a workflow connects untrusted information to tools that can change business state. Name the affected workflow, the consequence of a wrong decision, and the owner who can act on the result. The initial discussion should establish which evidence is already available and which uncertainty is material enough to investigate.

The specific failure hypothesis is that a threat list focuses on the model while overlooking the connector that actually commits changes. It remains a hypothesis until examined in the deployment. Existing internal tests or operating records may already resolve it; identify the uncovered question before commissioning additional work.

Mechanism

Threat models connect attacker influence to business effect

Relationships within agent security

Source or conditionControlled input
Declare what the hypothetical attacker can change.
Crossing or choiceTrust crossing
Locate where external content gains influence.
Consequence or decisionConsequential action
Tie the hypothesis to an asset and control.
Observe the control and resulting business effect. Each relationship needs evidence in the actual scope.

What this specialty examines

Threat modelling should start with assets and consequential actions, then identify who can influence the path to them. Separate user instructions, external content, peer agents and compromised dependencies. Each gives a hypothetical attacker a different level of control.

The review should map trust boundaries and attacker-controlled inputs to consequential actions and proposed controls. Identify the normal route, relevant exceptions and the authoritative record of the outcome. Inspect the relevant business-system record, enforcement event or process evidence alongside any reported status. Keep incomplete cases visible, with an explanation of what their absence prevents the owner from concluding.

The tradeoff to make explicit

For this decision, broad threat coverage aids discovery while a specific deployment is needed to make testable claims. Compare the available design or operating choices against the same intended outcome. If one option removes a required control or changes the task boundary, describe that difference before treating its apparent simplicity, speed or lower effort as an improvement. The decision should expose the tradeoff instead of burying it in an aggregate score.

Preparation

Connect the working inputs to the boundary being examined

Evidence packet for controlled input

Illustrative working format

BasisData-flow diagrams
Match this description to the deployed or proposed boundary being examined.
ContextActor assumptions
Separate observed inputs from assumptions and retain the comparison boundary.
ObservationAssets
Name the custodian and the conditions requiring a refreshed version.
DependencyAction privileges
Identify provenance and the specific decision this material can inform.
Illustrative preparation record. Availability, revision, and handling need confirmation.

Start with data-flow diagrams, actor assumptions, assets and action privileges. Identify the owner, revision and operating context of each source. Sanitised examples may support initial design, but differences from actual conditions should remain explicit. Where direct access is unsuitable, a customer-controlled evidence or execution route may be considered, with agreed handling and retention.

The scope also needs the security decision owner, agreed expectations for a representative threat scenario, and the authority to inspect the relevant boundary. Specify what evidence would distinguish the failure in which a threat list focuses on the model while overlooking the connector that actually commits changes from acceptable behaviour. If stakeholders disagree about intended behaviour, resolve the interpretation before classifying the observed difference as a defect.

How the work proceeds

Carry controlled input through to an accountable finding

From the scoped material to a decision record

  1. Establish conditionsData-flow diagrams

    Define the hypothetical attacker’s starting influence.

  2. Examine the relationshipTrust crossing

    Apply the scoped check to consequential action.

  3. Retain the findingAgent threat model

    Record the evidence, open dependencies, and the owner’s next decision.

Proposed sequence; resolve interpretation or access gaps before relying on the next step.

Define the controlled threat scenario

Review data-flow diagrams, actor assumptions, assets and action privileges with security architects and agent product owners. State what the hypothetical attacker controls, the protected action and the systems included in the authorised exercise. Prepare synthetic records and harmless markers where appropriate, with stopping conditions for unexpected effects. The preparation output is a scenario brief that distinguishes an influence attempt from a prohibited action actually reaching a business resource.

Challenge the path and review effects

Trace influence through planning and tools to the resource that commits an action. Prioritise hypotheses using consequence, reachability and existing controls. Keep evidence-backed observations distinct from plausible paths that still require validation.

Select a control response and retest

Build an agent threat model with ranked hypotheses and validation priorities from the reviewed observations. Separate agent behaviour, application enforcement and the resulting resource state so the recommended change reaches the responsible layer. Choose a bounded retest that reproduces the original condition and checks legitimate behaviour after remediation. If the required investigation exceeds the agreed technique or expertise, retain the unresolved path and identify the qualified review needed.

Indicative artefact

Make the agent threat model usable by its owner

Agent threat model

Illustrative working format

Annotated fieldAgent threat model
Identify the assessed workflow, revision, and decision boundary.
Annotated fieldRanked hypotheses
Link supporting and contrary observations to inspectable evidence.
Annotated fieldValidation priorities
Name the missing evidence, responsible owner, and condition for revisiting the decision.
Illustrative artefact structure, not a completed engagement or an observed result.

A useful decision artefact is an agent threat model with ranked hypotheses and validation priorities. Supporting material may include selected cases, configuration or process records, evidence references, a disagreement log and prioritised actions. Retain the case selection rationale and material disagreements alongside the result; otherwise the result may be mistaken for complete coverage.

Illustrative scenario

Examine consequential action through a constructed exception

A constructed case and its decision paths

Illustrative scenario

  1. SituationControlled input

    An untrusted document tries to redirect an agent towards an unauthorised action.

  2. Question to examineTrust crossing

    Trace the document's influence from ingestion to requested tool action.

  3. Conditional next actionConsequential action

    Choose controls at the consequential trust crossing identified by the case.

Fictional example. No executed test, customer engagement, or measured outcome is implied.

Consider a fictional situation: a supplier document influences an agent that can amend purchasing records and send external messages. The scenario is illustrative, with no client engagement, executed test or measured improvement implied.

For the supplier document, map its ingestion and the agent’s purchasing authority. The first useful question is whether document content can change action parameters or approval expectations. That path can guide a bounded test without assuming an attack has occurred.

Scope and next action

Keep the conclusion within the agreed threat scenario

Confirm fit before the next commitment

Included boundaryControlled input
This review is limited to the agreed threat scenario.
PrerequisiteData-flow diagrams
The relevant owner must confirm access and the intended operating conditions.
Claim limitAgent threat model
A bounded challenge cannot establish exhaustive protection.
Bring the triggering change and existing evidence; resolve the dependency that could alter the decision.

Authorised scenario testing is not a claim of exhaustive penetration testing or guaranteed protection. Intrusive security work requires qualified specialists and a separately agreed scope. Any engagement depends on an agreed scope, qualified team, authorised access and confirmed availability. Bring the affected workflow, its current configuration or process, the triggering change and the evidence already available. The immediate scope question is prompted by the situation in which a workflow connects untrusted information to tools that can change business state; confirm which missing observation would change the owner’s next action.

Ways to engage

Related engagement scopes

Agent Release Readiness Review

The Agent Release Readiness Review is a fixed-scope, three-week independent test of one AI agent workflow that shows a release board what the agent does, what it is allowed to do and what happens when it is wrong, and ends in a signed recommendation.

3 weeksRead about this offer

AI Agent Runtime Security Review

Make every delegation explicit. Examine who the agent acts for, where it can execute and which tools it can reach when instructions alone are not enough.

Read about this offer

AI Agent Red Team

Challenge the boundaries that matter to your workflow. Test agreed attack scenarios against agent instructions, tools and data access.

Read about this offer

Continue exploring

Related insights and resources

Agent security6 MIN READ

The permission boundary behind every agent action

An agent can only be evaluated in the context of what its tools allow. Examine identity, delegated authority and action boundaries before you put a workflow in motion.

Whitepaper5 MIN READ

Before your agent can act, define its authority

A working paper for the moment an AI system moves from advice to action. Map identity, permissions, approvals and recovery into an authority model your team can review and test.

Agent security6 MIN READ

Running an AI-agent incident tabletop exercise

A useful tabletop follows one plausible agent failure through detection, containment, evidence, business recovery and accountable return to service.

A little more clarity

Questions to consider.

Start with the questions that shape a better decision.

Should the threat model cover only the language model?

No. Cover the deployed workflow, including data sources, identity, memory, orchestration, tools and business systems. The model is one component in the path from influence to consequential action.

Which assets and actions belong in the threat model?

Start with data-flow diagrams, actor assumptions, assets and action privileges. Record their revision, owner and relationship to the intended workflow. Missing context should be identified before interpreting outcomes; a sanitised fixture is useful only when its relevant differences from the operating case are understood.

How should attacker influence be traced to business effects?

Map trust boundaries and attacker-controlled inputs to consequential actions and proposed controls. State the hypothetical attacker's control and inspect attempted actions, tool enforcement and actual effects separately. A refusal, a blocked call and a successful prohibited action are different results. Preserve unsuccessful challenges as well as adverse findings.

What should a prioritised threat model contain?

The proposed output is an agent threat model with ranked hypotheses and validation priorities. Include the selected conditions, supporting observations, unresolved differences and next decision. A reviewer should be able to follow each important conclusion back to its evidence and understand the limits without depending on an undocumented explanation from the author.

When should a hypothesis become a test priority?

Prioritise hypotheses that connect credible attacker influence to consequential actions and lack adequate control evidence. Record reachability and assumptions. A dramatic scenario without a plausible deployment path may deserve less immediate testing.

What if existing evidence already answers the question?

Review its scope, provenance and configuration before proposing new work. Existing work from security architects and agent product owners may already be sufficient. In that case, capture the decision, unresolved limitations and the change that would require another review. Additional assessment should address a real gap.

What would make the findings inconclusive?

Missing data-flow diagrams, actor assumptions, assets and action privileges can prevent a reliable interpretation, depending on the case. Record the specific missing observation and its implication. The owner can supply evidence, narrow the claim or defer the dependent decision; an incomplete case should not silently be counted as a pass.

Start with the decision you need to make.

Tell us about the agent workflow you want reviewed, and where you need greater clarity. Scope and availability are confirmed together.

Book a 30 minute readiness call Get the agent release checklist

Connected disciplines

Secure agent architecture assessment

Assess where security controls should sit in an agent architecture. The review connects component diagrams, identity flows, policy enforcement and data paths to a bounded review of agent security, giving the security decision owner a clearer basis for the next action.

Prompt-injection testing

Test whether direct instructions can redirect an agent beyond its approved task. The review connects instruction hierarchy, protected behaviours, synthetic prompts and tool audit records to a bounded review of agent security, giving the security decision owner a clearer basis for the next action.

Indirect prompt-injection assessment

Examine instructions hidden inside information an agent consumes. The review connects content provenance, retrieval paths, trust labels and consequential action rules to a bounded review of agent security, giving the security decision owner a clearer basis for the next action.