Quick answer
Breach and attack simulation (BAS) is a category of security tools that automatically and repeatedly run safe versions of known attack techniques against your environment, to check whether security controls block or detect them. BAS gives continuous measurement of control coverage, but does not replace human-led testing.
On this page
How BAS Works
Agents or virtual machines in your environment execute libraries of attack techniques, usually mapped to MITRE ATT&CK: simulated malware, lateral movement, data exfiltration, email attacks and more. The platform records which were blocked, which were detected and which went unnoticed, and repeats the tests on a schedule so you can see when coverage drifts.
What BAS Does Well
- Continuous checking that controls still work after changes.
- Broad, repeatable coverage of known techniques.
- Measuring the effect of new detection rules or security tools.
Where BAS Falls Short
BAS runs scripted techniques. It does not chain weaknesses creatively, adapt to your environment, exploit business logic, or think like a determined human. It cannot tell you whether a real attacker would find a way around the controls it tests. That is the job of penetration testing, purple teaming and red teaming.
Combining BAS With Human Testing
The strongest programs use BAS for continuous coverage of known techniques and human operators for depth: adversary simulation, red team engagements and investigation of what BAS finds. Our adversary simulation service combines both.
Regulatory and Framework Context
For Australian organisations, BAS results can provide ongoing evidence for APRA CPS 234 control testing and Essential Eight maturity between formal assessments.
Common questions
Is BAS the same as penetration testing?
No. BAS automatically checks known techniques against controls. Penetration testing uses human expertise to find and chain vulnerabilities in your specific environment.
Is breach and attack simulation safe?
Reputable BAS platforms use safe, non-destructive versions of techniques, but should still be scoped and monitored carefully, especially in production.
Change log
- 13 November 2026First published.