Always-On Agentic Pentesting for Engineering Teams

Barrion runs continuous, agentic AI penetration tests of web applications and APIs. AI agents test your app the way an attacker would, safely, on a schedule or on demand. Findings are checked against the live app before they're reported, and tracked across runs as new, still open, resolved or regressed. From Standard level up, a security engineer reviews each report. Barrion AB is based in Gothenburg, Sweden. A free passive scan is the quickest way to start.

AI pentests send real test requests, so they're rate-limited and non-destructive, and you approve the exact scope before a single request goes out. You can point them at staging too. The free passive scan only reads your live app. It never submits forms, brute-forces endpoints or touches anything that changes state, so it's safe to run against production.

AI pentests look for the issues an attacker could exploit, like SQL injection, cross-site scripting and broken access control. Findings are checked against your live app, and confirmed ones come with the request and response that prove them. Anything we couldn't confirm is clearly marked and capped in severity. The passive scan catches misconfigurations in TLS and HTTPS, security headers, cookie flags, CORS policy, DNS records, email authentication (SPF, DKIM, DMARC) and network exposure.

Barrion covers two surfaces. AI pentesting is the active part: specialist agents chain requests to find exploitable flaws such as SQL injection, cross-site scripting and broken access control. Findings are checked against your live app, and confirmed ones come with the request and response that prove them. The passive scan is read-only and safe to point at production. On paid plans it runs 35+ checks (18 on the free plan): transport security (HTTPS, HSTS, TLS version, cipher suites, certificate expiry, hostname and chain validity, OCSP stapling), HTTP response headers (Referrer-Policy, Permissions-Policy, X-Content-Type-Options, Content-Type, server information disclosure), CSP and framing (Content-Security-Policy, bypass detection, Trusted Types, X-Frame-Options), cross-origin policy (COOP, COEP, CORP and the full CORS header set), cookie flags and anti-CSRF tokens, mixed content, vulnerable JavaScript libraries, DNS records including DNSSEC and CAA, email authentication (SPF, DKIM, DMARC), open ports and subdomain takeover. Findings from both are ranked by severity and come with step-by-step remediation.

The report maps every finding to OWASP WSTG, a security engineer reviews it from Standard up and signs it off from Deep up, and a free retest shows what you fixed. Teams use it as evidence for SOC 2, ISO 27001 and NIS2 work. Whether it's accepted is up to your auditor.

Continuously or on demand. On the Business plan you save a pentest and it reruns daily, weekly, monthly, quarterly, every six months, yearly or on your own rhythm. On any plan you can start a pentest or a passive scan on demand. A passive scan also runs on a schedule and alerts you when something new shows up.

Yes. Solo developers often start with the free passive scan and a single pentest. Teams with several apps run pentests on a schedule, and it fits alongside the tools you already run.

Passive scans are read-only: they only look at what your app already exposes publicly. For AI pentests we store the findings and the evidence behind them, the requests and responses that confirm each issue, so you can review and reproduce them. Test credentials you add are encrypted. Data is stored and hosted in Sweden and AI processing runs in the EU. Our trust page lists every subprocessor. Pentests are rate-limited and only run inside the scope you approve.

You can cancel anytime in the dashboard, and paid plans carry a 14-day refund window from the first charge. If something isn't right, contact us and we'll make it work for your team.

Barrion's pentest reports come as PDF, XLSX and JSON, mapped to all 97 OWASP WSTG test cases, with engineer review from Standard up, sign-off from Deep up and a free retest after fixes. Teams use them as evidence that supports SOC 2, ISO 27001, PCI DSS and NIS2 work. Whether a report is accepted is up to your auditor or customer.