StartCloud's cyber security team runs AI-assisted penetration tests (ethical hacking) for Australian businesses, most with 5 to 50 staff: authorised, simulated attacks on internet-facing systems, internal networks, web applications and APIs, Microsoft 365 and Entra ID, and Azure and AWS cloud configuration, plus optional phishing simulations. Frontier AI models available only to verified security teams help analyse and validate findings, and every finding is checked by a person. Testing happens only against targets the client approves and signs off in writing, following methodologies such as OWASP and PTES. Data analysed with AI is processed by the AI provider, which retains records to monitor for misuse. Each test includes a report with risk ratings and fixes, help remediating, and a retest, at a fixed price quoted after scoping. Based in Balcatta, Perth.
Penetration testing
AI-assisted penetration testing, led by our security team
Our security team tests your systems the way a real attacker would, with your written sign-off and only against the systems you approve. Frontier AI helps us analyse and confirm what we find, so you get more confirmed findings, sooner, and a clear list of what to fix.
Every finding is checked by a person before it reaches you. Then we help you fix it, and retest.
5.0on GoogleMicrosoft Solutions Partner
Quick answer
Penetration testing, also called ethical hacking, is an authorised, simulated attack on your systems to find weaknesses before a criminal does. StartCloud's cyber security team runs penetration tests for Australian businesses, most with 5 to 50 staff, covering internet-facing systems, internal networks, web applications, Microsoft 365 and cloud configuration, and phishing simulations. We use frontier AI models, available only to verified security teams, to help analyse and validate findings, and every finding is checked by a person. Testing happens only against targets you approve and sign off in writing. You get a report with risk ratings and fixes, help remediating, and a retest. Each test is a fixed price, quoted after scoping.
Is this you?
AI is helping attackers find weak spots faster. Your insurer, a client or a tender wants to know when you last checked yours.
01
More found, sooner
AI helps our team analyse and confirm weaknesses faster than a manual test alone, so more gets covered in the same engagement.
02
Only what's real
Every finding is validated and checked by a person, so you're not chasing false alarms.
03
Fixed, then checked
We help close the gaps, then retest to confirm the fixes worked.
04
Proof when you're asked
A report and retest results you can show an insurer, client or auditor.
What's included
An AI-assisted penetration test
Scoped to you and signed off by you, run by our security team with AI analysis, then the fixes and a retest.
- Format
- Scoped, fixed price, signed off by you
- Covers
- External, internal, web apps, Microsoft 365, cloud, phishing
- Outcome
- Weak spots found, confirmed, fixed and retested
Fixed price, quoted after scoping
- A scoping call, and written sign-off on targets, timing and rules of engagement
- Testing only against the systems you approve
- External, internal network, web app, Microsoft 365 and cloud testing as needed
- AI-assisted analysis and validation of findings, reviewed by our team
- Optional phishing and social engineering simulations
- A report with risk ratings and fixes, help remediating, and a retest
How it works
- 01
Scope and sign-off
We agree targets, timing and rules of engagement, and you sign off in writing. Nothing outside that list is touched.
- 02
Test
Our security team tests using recognised methodologies such as OWASP and PTES, with AI helping analyse what we find.
- 03
Validate and report
Every finding confirmed and checked by a person, then written up for management and for whoever fixes it.
- 04
Fix and retest
We help close the gaps, then retest to confirm they're closed.
The tools
What we test
External and internet-facing
Websites, remote access, VPNs, email gateways and anything else reachable from the internet.
Internal network
Servers, workstations and network segmentation, and what an attacker could reach from one laptop.
Web applications and APIs
Login, access control and data exposure flaws, tested against OWASP guidance.
Microsoft 365 and Entra ID
Sign-in settings, admin access, sharing and conditional access gaps.
Cloud environments
Azure and AWS configuration, permissions and exposed services.
Phishing and social engineering
Simulated phishing and pretext calls to see how your team responds.
Common questions
Questions about penetration testing
What is AI-assisted penetration testing?
A penetration test run by our security team, using frontier AI models to help analyse and validate the weaknesses we find. The AI speeds up the analysis. People decide what's tested, check every finding and write the report.
Is it safe to use AI in a test of our systems?
Testing only happens against systems you've approved and signed off in writing, within agreed times and rules. Our team runs and supervises the work. The AI models we use are only available to verified security teams, and they have built-in blocks on harmful actions.
Where does our data go?
Technical details we analyse with AI are processed by our AI provider, which keeps records so it can monitor for misuse. We only use what the test needs, and we set this out in the scope before you sign off. If some data can't leave your environment, tell us and we'll test those systems without AI analysis.
Does AI replace human testers?
No. AI makes our team faster and more thorough, but people decide what to test, judge what a finding means for your business, and stand behind the report.
Why does AI change penetration testing?
Because attackers have AI too. Weak spots that used to take a long time to find can now be found much faster, so testing once a year and hoping for the best carries more risk than it used to. Using AI on the defending side helps even that up.
Is ethical hacking the same as penetration testing?
Near enough. Ethical hacking is the broader term for authorised hacking that improves security. A penetration test is a scoped, time-boxed ethical hacking engagement with a report at the end.
What's the difference between a penetration test and a vulnerability scan?
A vulnerability scan is automated and runs regularly to find known weaknesses. A penetration test is hands-on: people try to break in and show what an attacker could do with what they find. Most businesses do best with scanning all year and a penetration test at least once a year.
How often should we have a penetration test?
At least once a year, and after any big change, like a new website, a move to the cloud or a new office network. Some insurers, clients and frameworks ask for it on a set schedule.
Does the Essential Eight require penetration testing?
No. Penetration testing isn't one of the eight controls. It is still one of the best ways to check your controls are working, and showing that controls work is exactly what ASD Essentials, the Essential Eight's successor, puts more weight on.
How much does a penetration test cost?
It depends on what's being tested and how deep the test goes. Each test is a fixed price, quoted after a scoping call.
Related services
Further reading
- Cyber insurance: what it covers. And the controls insurers ask about
- The Essential Eight is changing. What ASD Essentials means for your business
- How ransomware gets into a business. The paths a test usually finds
- Security awareness training. When the phishing test shows gaps in habits
- Cyber security services in Perth. Everything we do, for Perth businesses
Ready when you are
Let's go beyond.
Ground Floor West, 6 Gibberd Road, Balcatta WA 6021