AI-assisted pentesting by Excession Labs
AI pentesting that shows its evidence
Incursion, our testing agent, works through your web application and API as each user role you give it. It may only report a finding that links to proof, and Excession Labs checks that proof before the finding reaches your report.
For ordinary web apps and APIs. Testing starts only after written scope and authorization.
| Candidate | Check | Verdict |
|---|---|---|
buyer_A reads /v2/orders/ord_test_B1 and gets 200 | Nonexistent order returns 404. buyer_B's own session shows the same order and total. | Confirmed, High |
/v2/me/orders?user=buyer_B returns 200 | The response holds buyer_A's own orders. The API ignores the parameter. | Dropped |
| Refund endpoint accepts a negative amount | State-changing request: rule set to ask. Approved by Excession Labs, run on a test order. | Confirmed, Medium |
Password guessing on /login | Rule set to deny for this scope. Not run. | Out of scope |
Every confirmed finding links to its requests, responses and controls. See the full example
How a test runs#
Incursion splits the engagement into tasks and runs them in parallel. Tool rules decide what it may do, and a task ends only when its findings link to evidence.
- Scope and authorize
Hosts, roles, test windows and off-limits actions, signed by the system owner.
- Set the rules
Each tool runs, asks for approval, or is blocked. Costly and state-changing actions can wait for a person.
- Test in parallel
Workers test roles, endpoints and flows at the same time, and keep notes.
- Prove or drop
Controls and read-back from a second account decide what is real.
- Review and report
Excession Labs checks each finding's evidence and writes the report. Then we retest your fixes.
What it tests#
- Web applications with signed-in user flows, for each role you provide
- REST and GraphQL APIs, with or without documentation
- Object-level and function-level authorization between users and tenants
- Sessions, login, password reset and account flows
- Input handling: injection, stored scripts, file uploads
- Admin panels and internal tools exposed to the web
Out of scope here: internal networks, physical and social engineering tests, and testing of AI models themselves. For LLM apps and agents see AI red teaming.
Why each finding needs evidence#
A language model can read a 200 response as a breach when the server simply ignored a parameter. Incursion records the request, the response and the controls for each candidate, so a person can check the claim. In the example above, one candidate looked like a leak until the control showed that the API returned the caller's own orders. We dropped it.
Who does what#
| Step | Incursion | Excession Labs |
|---|---|---|
| Scope and authorization | Agrees it with you in writing | |
| Tool rules | Follows them | Sets them for your scope |
| Mapping and testing | Runs tasks in parallel, as each role | Approves actions that need approval |
| Evidence | Links each finding to requests, responses and controls | Checks the evidence |
| Report | Drafts findings with reproduction steps | Writes and signs off the report |
| Retest | Reruns the proof against your fix | Confirms the result |
You decide what is off limits#
We agree in writing which actions must never run, such as deleting data, sending real email or charging cards. Destructive proofs use only test-owned accounts and records. Tests that cost money, like email sending or paid third-party APIs, are bounded, and the report states what they cost.
After each release#
Once your application is mapped and the test accounts exist, testing again after a meaningful release is faster to set up. Testing after releases
Guides to AI-assisted testing#
Questions
What is AI penetration testing?
It is penetration testing in which an AI agent runs much of the testing work: mapping the application, sending requests as different users, comparing responses and collecting evidence. In our service the agent is Incursion. Excession Labs scopes the work, checks the evidence for each finding and writes the report.
Is this the same as testing an AI chatbot?
No. This site uses AI to test ordinary web applications and APIs. If you need your own LLM application, RAG system or agent tested for prompt injection and data leakage, see airedteamingservices.com.
Does the agent run without supervision?
It runs under tool rules we set for your scope. Each tool either runs on its own, waits for approval, or is blocked. State-changing and costly actions can require approval. Excession Labs checks the evidence before anything goes in your report. See safety and control.
How is this different from a vulnerability scanner?
A scanner sends known payloads and reports pattern matches. The agent works through the application as different users, follows multi-step flows and must link each finding to evidence. Read AI pentest agents vs DAST scanners.
How much does it cost?
We quote after scoping, because the work depends on the number of applications, roles, API endpoints and environments. You can start with a scoped pilot. See pricing.
Run it on your application
Start with a scoped pilot. Incursion tests, a person checks each finding, and you get a report with evidence.