Australian-hosted by default

The platform is hosted in Australia on infrastructure operated by Extranet Systems. Customer data is stored in Australia.

Works inside segmented networks

The on-premise agent connects outbound-only over TLS 443 with 0 inbound firewall rules, and survives restrictive corporate proxies.

Identity and privilege testing

Active Directory enumeration, credential testing and password-policy auditing show where standard accounts reach administrative control.

Proof, not guesswork

Safe automated exploitation is gated by per-tenant Rules of Engagement, with scope enforcement and a full evidence trail.

Compliance-mapped reporting

Findings map to 8 compliance reporting frameworks plus CIS Benchmarks, and export as PDF, CSV and JSON/API.

Coverage between annual tests

Scheduled scans, 7-day asset re-verification and drift detection keep the picture current after an assessment has been signed off.

What makes government environments different

Few sectors carry as much technical history. A single agency can run a modern cloud-hosted citizen portal, a decade-old line-of-business application that nobody dares patch during business hours, an on-premise directory that predates the current operating model, and a network shared with entities that have since been split off or merged in.

That estate is also unusually exposed. Citizen-facing services are published deliberately, discoverable by design, and cannot be taken offline for convenience. Machinery-of-government changes move systems, staff and trust relationships between entities faster than documentation is updated. Delivery is frequently outsourced, so the real attack surface extends into suppliers and shared-services providers.

The result is a gap between what the asset register says and what is actually reachable. Testing closes that gap by attacking the estate the way an adversary would, then proving which paths work and which are theoretical.

The policy drivers behind government testing

Public sector security programmes are shaped by policy rather than by a single regulator. The Essential Eight mitigation strategies, the Information Security Manual and the Protective Security Policy Framework set direction at Commonwealth level, while states, territories and local councils layer their own obligations on top. The Privacy Act 1988 and the Notifiable Data Breaches scheme apply where personal information is involved.

Penetration testing does not satisfy any of these on its own. What it does is supply the technical evidence that policy work is usually missing: whether the controls described on paper hold up when someone actually tries them.

Policy driverWhat exploit-validated testing evidences
Essential Eight mitigation strategiesPatch currency on internet-facing and internal systems, and whether administrative privilege restrictions hold when a standard account is compromised
Information Security Manual and PSPF directionTechnical control effectiveness across the perimeter, internal network, applications and cloud configuration
Privacy Act 1988 and the NDB schemeWhich paths lead to systems holding personal information, and how far an attacker gets before detection
Assurance and audit programmesRepeatable, dated findings with evidence, a 365-day audit log and retention of up to 7 years on Enterprise plans
Supplier and procurement questionnairesCurrent findings and remediation status you can attach to a response instead of reissuing last year's report

An honest note on the Essential Eight

PentestOps reports map findings to 8 compliance reporting frameworks (OWASP Top 10, PCI DSS v4.0, NIST 800-53, SOC 2, HIPAA, GDPR, ISO 27001, SMB1001) plus CIS Benchmarks for AWS, Azure, GCP and Kubernetes. The Essential Eight is not one of those mappings, and testing is not a maturity assessment.

Testing produces evidence relevant to several of the strategies, particularly patching and the practical limits of administrative privilege. It says nothing about backup regimes or application control policy, which need separate review. Treat the two as complementary and you will get more value from both.

Regulatory landscape

Public sector obligations arrive from policy and legislation rather than from a single regulator, and they ask for different things. The ACSC Essential Eight sets a control baseline with maturity levels, the Information Security Manual supplies the control catalogue, IRAP is how an independent assessment against that catalogue is performed, the Security of Critical Infrastructure Act places duties on responsible entities for critical infrastructure assets, and the Protective Security Policy Framework frames protective security across the entity.

Two boundaries before the table. PentestOps does not make an agency compliant with any of these, and it cannot assess Essential Eight maturity. PentestOps is also not IRAP assessed and holds no government accreditation; where a hosting or assessment requirement applies to us as a vendor, raise it early and we will tell you plainly whether we meet it.

ObligationWhat it expects of youEvidence a security test provides
ACSC Essential Eight and its maturity levelsMitigation strategies implemented to a target maturity level, with the level verified rather than assumedTechnical evidence for part of the picture: patch currency on internet-facing and internal systems, and whether restrictions on administrative privileges hold when a standard account is compromised. It is not a maturity assessment
Information Security Manual (ISM)Controls selected from the catalogue, implemented, and their effectiveness assessed rather than assertedEffectiveness evidence for the technical subset: perimeter exposure, internal segmentation, identity and privilege, and application and cloud configuration, dated per scan
IRAP assessmentAn independent IRAP assessor evaluates the system against the ISM and records the residual risk for the authorising officerValidated findings, evidence and remediation status the assessor can review as technical input. The assessment itself is the assessor's work, and PentestOps is not IRAP assessed
Security of Critical Infrastructure (SOCI) ActA risk management programme that identifies and mitigates cyber and information security hazards to a critical infrastructure asset, and is kept currentEvidence on the cyber hazard side: which exposures are genuinely reachable across the asset in scope, re-tested as the environment changes, with drift detection in between
Protective Security Policy Framework (PSPF)Security risk managed across the entity, with maturity reported annually against the core requirementsThe dated technical findings behind the technology statements: what was exploitable, what has since been closed, and scan comparison between reporting cycles

Sovereignty, hosting and the questions procurement will ask

Where the data lives is usually the first question and the one that stops most evaluations. The platform is hosted in Australia on infrastructure operated by Extranet Systems, and customer data is stored in Australia. Specific data-residency arrangements are available to Enterprise customers on request, along with an on-premise deployment option where an agency needs the platform inside its own boundary.

Every customer runs in an isolated Kubernetes namespace with its own dedicated database, so there is no shared tenant data store. Data is encrypted at rest, traffic is protected with TLS in transit, and credentials and signing keys are held in a dedicated secrets vault and injected at runtime. Role-based access control and multi-factor authentication are available on every account and required for administrators. Full detail sits in our Trust Centre.

On the vendor side, Extranet Systems Pty Ltd is ISO/IEC 27001:2022 certified, independently audited by Atom Assurances, and the platform is built to SOC 2-aligned controls with a SOC 2 attestation on our roadmap. We do not claim assessments we do not hold. If your agency requires a specific hosting arrangement or third-party assessment, raise it early and we will tell you plainly whether we meet it.

How PentestOps tests a government estate

Assessments follow a seven-phase methodology aligned to PTES, OWASP Web Security Testing Guide v4.2 and NIST SP 800-115, and aligned to CREST testing methodology guidance. Every scan and every exploitation attempt is tied to signed Rules of Engagement, with scope enforcement that automatically stops activity outside authorised assets.

  • Perimeter. 24+ external recon modules map what the public internet can see, including services published years ago and never decommissioned. See external network testing.
  • Internal network. The on-premise agent runs 18+ internal modules natively on the LAN instead of tunnelling each packet out to a remote scanner. See internal network testing.
  • Identity. Active Directory testing covers enumeration, credential testing and password-policy auditing, then validates the chain from one weak credential to a full pivot.
  • Citizen-facing applications. Web and API testing across OWASP Top 10 and API Top 10 risks, covering REST and GraphQL, with a live finding stream while the scan runs.
  • Cloud and containers. 800+ automated checks across AWS, Azure, GCP and M365 with CIS Benchmark coverage, plus agentless Kubernetes auditing through a read-only kubeconfig.
  • Prioritisation. CVE correlation, CVSS v3.1 scoring, exploit-availability indicators and CISA KEV prioritisation, so a small team works the findings that actually matter first.

Typical use cases

Government engagements tend to fall into a handful of recognisable shapes. The common thread is that someone needs defensible evidence quickly, without a procurement cycle for every question.

  • A council with a two-person IT team that owns dozens of internet-facing services and cannot fund a manual test for each one.
  • Pre-go-live assurance for a new citizen-facing service, run against staging and then re-run after fixes to confirm closure.
  • Coverage between annual assessments, where an independent test happens once a year but the environment changes weekly. This is the case for continuous testing.
  • Post-change validation after machinery-of-government transfers, when systems, trusts and administrative accounts move between entities.
  • Shared-services providers testing many entities from one platform, with isolation between each tenant.
  • Supplier assurance, where a delivery partner's environment is in scope and properly authorised under the Rules of Engagement.

Why agencies choose PentestOps

  • Australian-built and operated, with customer data stored in Australia and Enterprise data-residency options.
  • Extranet Systems Pty Ltd is ISO/IEC 27001:2022 certified, independently audited by Atom Assurances.
  • Evidence-first: safe automated exploitation proves impact rather than handing over a list of theoretical severities.
  • No inbound firewall rules, no VPN and no jump host to test internal networks. The agent deploys in about 5 minutes.
  • Asset-wise pricing: you pay for the assets in scope, and scans against them are unlimited within fair use.
  • Self-hosted AI by default, so scan data is analysed on infrastructure Extranet Systems operates rather than being sent to third-party model providers. External providers are opt-in per tenant.

Frequently Asked Questions

Does PentestOps assess our Essential Eight maturity?

No. Reports map findings to 8 compliance reporting frameworks (OWASP Top 10, PCI DSS v4.0, NIST 800-53, SOC 2, HIPAA, GDPR, ISO 27001, SMB1001) plus CIS Benchmarks, and the Essential Eight is not one of those mappings. Testing does produce technical evidence relevant to several strategies, especially patch currency and whether administrative privilege restrictions hold in practice. Use it alongside a maturity assessment, not instead of one.

Where is our data stored?

The platform is hosted in Australia on infrastructure operated by Extranet Systems. Customer data is stored in Australia. Specific data-residency arrangements are available to Enterprise customers on request, and Enterprise also includes an on-premise deployment option.

Can it test networks that allow no inbound connections?

Yes. The on-premise agent connects outbound-only over TLS 443 through a reverse tunnel, so 0 inbound firewall rules are required. It has no inbound listeners, supports HTTP CONNECT through restrictive corporate proxies, and a single host can cover multiple subnets. See the agent for the full architecture.

Is automated exploitation safe to run against live citizen services?

Exploitation only runs where you authorise it. Every scan and exploit is gated by per-tenant Rules of Engagement with scope enforcement that automatically stops activity outside authorised assets, and a strategy engine picks the safest technique that proves the finding. Stealth, Balanced and Aggressive scan profiles let you dial the intensity down for production, and scans can be scheduled outside service hours.

Does this replace an independent penetration test?

No, and we will not pretend otherwise. Automation runs the repetitive, high-volume testing so your people and your independent testers can focus on the creative work that automation handles poorly, such as business logic flaws and process abuse. Most agencies keep their periodic independent assessment and use PentestOps for continuous coverage in between.

How is it priced for an agency or council with a fixed budget?

Pricing is asset-wise. One asset is one item the platform can scan or monitor: a public IP, hostname, web application, internal subnet target, cloud account or Kubernetes cluster. Scans against an asset are unlimited within fair use, so testing more often costs nothing extra. Live plan detail is on pricing.

How long are findings and audit records kept?

Assessment findings are retained for 1 year on Starter, 3 years on Professional and up to 7 years on Enterprise, then automatically deleted. Audit logs are retained for 365 days. Exported reports, signed authorisation records and legal documents are kept and are not removed by the retention process, and findings can be placed on legal hold.

Can we test a supplier's or shared-services provider's systems?

Only where you have authority to do so. Scope is enforced against signed Rules of Engagement, so assets belonging to a third party must be authorised before they can be tested. If a shared-services provider runs the environment, the cleanest arrangement is usually for the provider to hold their own tenant and share results with each entity.

Prove what an attacker could reach

Start with your perimeter and expand into the internal network when you are ready. All paid plans start with a 7-day free trial. A card is required to start your trial and is only charged after the trial ends, unless you cancel first.