In summary
- Parabellum Cybersecurity has become one of just 13 firms worldwide, and three in Asia Pacific, to gain CREST International's AI-Enabled Penetration Testing accreditation.
- The standard independently verifies that AI is used responsibly, securely and under qualified human direction, sitting within CREST's binding complaints and discipline framework.
- CREST research shows 69% of penetration testing providers already use AI, with 85% expecting clients to demand greater transparency about its use.
Australian cybersecurity consultancy Parabellum Cybersecurity has been accredited by CREST International for AI-Enabled Penetration Testing, joining a small group of 13 firms worldwide to have passed the assessment since the standard was introduced in July 2026.
Parabellum is one of only three firms in Asia Pacific to hold the accreditation, which independently verifies that a provider uses artificial intelligence in its penetration testing services responsibly, securely and under the direction of qualified human testers.
The accreditation builds on Parabellum's existing CREST Penetration Testing accreditation and sits within CREST's complaints and discipline framework - making it a binding standard rather than a voluntary commitment.
CREST developed the AI-Enabled Penetration Testing requirements through its AI Working Group to address what it said was a growing gap between AI adoption and AI governance in the sector. Accredited providers are assessed on how they govern their AI tools, how they protect client data, and how they ensure a qualified human directs and validates every finding.
CREST's own research found that 69% of penetration testing providers already use AI, 76% increased their use over the past year, and 85% expect clients to demand more transparency about how AI is used in engagements.
For Parabellum's clients - which the company said range from SMEs to critical infrastructure operators and government - the accreditation is intended to provide assurance that AI-assisted testing delivers broader coverage and speed without compromising data security or professional judgement.
"Our clients trust us with some of their most sensitive systems and data," said Martin Dybalski, director of Parabellum.
"AI helps our offensive security specialists go further and faster, but it never replaces their judgement, and client data is only ever handled under strict, contracted controls. This accreditation is independent proof that we hold ourselves to that standard."
Stuart Shanahan, director of technical services at Parabellum, said the accreditation reflected how the threat landscape itself was shifting.
"We're seeing more and more adversaries using AI. To test our clients' defences realistically, experienced security professionals need to wield it too - but in a way that is controlled, auditable and accountable," he said.
Nick Benson, chief executive of CREST, said the accreditation was designed to give buyers greater confidence in how AI is being used within security testing services.
"Human expertise and professional judgement remain central to delivery," he said.
"Parabellum's achievement demonstrates its commitment to these standards and to delivering high-quality, trusted cyber security services.”
The accreditation follows Parabellum's earlier commitment to the CREST AI Charter and complements its ISO 27001 certification.
In August, Parabellum launched Otis, a proprietary AI capability its consultants will use to deliver penetration testing engagements.




