Penetration Testing as a Service (PTaaS) Explained
Skip to content

Penetration Testing as a Service

Penetration testing as a service (PTaaS) is penetration testing bought as a standing subscription, with one agreed scope and tests that run through the term.

On this page

Findings are delivered through a platform as each one is confirmed, and fixes are retested inside the same agreement. PTaaS describes how testing is bought and delivered, so the people and tools doing the testing vary from one provider to the next.

This guide explains how the delivery model works, how it compares with a project test, how findings and retests are handled, how to choose among PTaaS providers, how pricing is usually structured, and how the results serve audits and a wider exposure management program.

Penetration Testing as a Service Defined

PTaaS brings together pieces that a project test keeps apart: a scope that stays open for the term, a platform where results are delivered and worked, and testing that repeats as the environment changes. The testing itself is still a penetration test, which ASD’s Information Security Manual describes as an assessment that plays out real-world scenarios to reach a set goal, for example gaining control of critical systems or data.

Buyers often meet related terms in the same proposal.

  • Penetration testing as a service: The phrase names a delivery model, in which testing is arranged as a subscription with a held scope, a results platform and retests included in the term.
  • Pentest as a service: This is the same model under a shorter name, and some providers also call it a penetration testing subscription.
  • Project penetration test: A project examines named systems during a booked window and ends when the report is accepted.
  • Continuous penetration testing: This describes the practice of testing whenever systems change, which a subscription is one way to buy.
  • Agentic penetration testing: This describes a method, in which AI agents plan and run tests, and a PTaaS provider may use it for part or all of its testing.
  • Scanning subscription: A recurring vulnerability scan is a different service, since NIST SP 800-53 describes penetration testing as going beyond automated vulnerability scanning.

PTaaS and Project Testing Compared

The agreement is the first difference between the models, and it shapes the rest. A project is bought once for a fixed set of systems and dates, while a subscription holds the scope and the relationship for the whole term.

Area Project penetration test PTaaS subscription
Agreement A statement of work for each engagement. One agreement for the term, with the scope kept current.
Timing A booked window, often once or twice a year. A baseline, then tests on change, on a rolling schedule and on request.
New systems Covered at the next engagement or through a contract variation. Added to the scope once ownership is confirmed.
Results One report when the engagement closes. Findings released one by one as they are confirmed.
Retest Arranged separately, often weeks later. Typically started by the fix itself and included in the term.
Record A report per engagement. One running record of scope, tests, findings and retests.
Assurance A dated report from a named tester. A period report generated from the record, plus any independent test bought alongside.

Public bodies still use the project model for depth across a whole organization. CISA describes its Risk and Vulnerability Assessment as a two-week penetration test of an entire organization, with one week of external testing and one week on the internal network.

A project keeps its value for depth, independence and a single dated opinion. A subscription is suited to systems that change between those dates, and many organizations run both.

How the Delivery Model Works

A subscription runs on a scope and a test policy agreed once at the start, and every later test follows them. The platform then holds the inventory, the test history and the findings for the term.

  • Scope: The subscriber names an owner for every domain, address range and application, and newly found assets are added once the subscriber confirms they are its own.
  • Test policy: Exclusions, permitted hours, request rate limits and excluded techniques are written once and apply to every test, whether scheduled, triggered or requested.
  • Approvals: Higher-impact tests, such as a sign-in attempt with a leaked credential or testing with accounts the subscriber supplies, wait for a named person to approve them.
  • Baseline: The first pass tests every in-scope system, so later results can be read against a known starting point.
  • Triggered tests: A newly recorded host, a release, a published advisory affecting an exposed product, or a leaked credential starts a test of the affected system.
  • Rolling schedule: Systems that see no change are visited again at a set interval, so none goes untested for the whole term.
  • Requested tests: The subscriber can name a system for testing, for example ahead of a launch.

The triggers and the records that reveal them are explained in our guide to continuous penetration testing. In a PTaaS agreement they are written into the contract, so the subscriber knows which events start a test and which wait for the schedule.

The platform is where both sides work. It typically shows the current scope, the tests that ran and why, each finding with its proof and status, and a way to request a test or report a fix.

Findings Delivered as Confirmed

A PTaaS provider releases each finding once it has been confirmed, without waiting for the end of a testing period. The owning team can start a fix while the change that caused the weakness is still fresh.

Each finding record names the affected asset, describes the weakness, holds the proof, and gives a severity and fix steps, as a finding in any penetration test report does. A subscription adds fields that a project report has no use for.

  • Trigger: The record names the event or schedule entry that started the test, which helps the team see which change introduced the weakness.
  • Status: The record shows the current state, such as open, fixed awaiting retest, closed or accepted.
  • History: The record lists each later step with its date, from the review and the ticket to the retest result.

Confirmation is the main quality control. A finding that arrives without proof sends an engineer to chase a possible false positive, so buyers should ask how each finding is confirmed and which ones a person reviews before release.

Retesting a Reported Fix

A retest runs the recorded proof again on the fixed system and notes whether the weakness is gone. PCI DSS requirement 11.4.4 expects penetration testing to be repeated to verify that exploitable vulnerabilities were corrected.

The PCI Security Standards Council’s Penetration Testing Guidance adds that a fix arriving long after the test may justify fresh testing, because the environment may have moved on in the meantime. It also notes that all changes made as a result of remediation should be retested, with risk assessment deciding whether a complete system retest is needed.

Within a subscription the retest usually follows this order.

  1. Report the fix. The owner marks the finding as fixed in the platform or in the linked ticket.
  2. Repeat the proof. The tester or agent sends the original request or repeats the original step against the same asset.
  3. Check for variants. The test tries close variations, such as a neighboring endpoint or a second parameter, since a fix often covers only the reported path.
  4. Record the result. A passed retest closes the finding with its date, and a failed one reopens it with new proof.
  5. Close the ticket. The ticket closes only on a passed retest, so a status change alone never counts as a fix.

Retest terms differ between providers. Buyers should confirm whether retests are unlimited, counted against an allowance, or charged, and how long after the original finding a retest is still offered.

Ticketing and Workflow Integration

A finding only reduces risk once someone fixes it, so a PTaaS platform is only as useful as its link to the teams that own the systems. Integrations send each finding into the queue each owning team already uses.

  • One ticket per problem: Each finding opens a single ticket, and repeat detections of the same weakness update it instead of opening another.
  • Field mapping: Asset, severity, owner and due date map to the ticket system’s own fields, so the finding sorts and reports like any other work item.
  • Routing: The asset owner recorded in the scope decides which team’s queue receives the ticket.
  • Two-way status: A status change in the ticket shows in the platform and the reverse, so neither side works from stale data.
  • Retest on fixed: Moving the ticket to fixed starts the retest, and the result is written back as a comment.
  • Proof stays protected: The ticket links to the proof in the platform, since ticket systems are often readable by many staff and the proof may show sensitive data.
  • SOC feed: Confirmed findings and test traffic details are shared with the security operations team, so analysts can separate test traffic from real attacks and check whether an alert fired.

Keeping proof behind a link follows from how standards treat test data. NIST SP 800-53 notes that penetration testing may expose information protected by laws or regulations to the people doing the testing, and agreements should say how that information is protected wherever it travels.

Choosing Among PTaaS Providers

PTaaS providers differ most in coverage, in how findings are confirmed, and in the commercial terms around scope changes and retests. The points below are worth settling in writing before the term starts.

  • Coverage by asset type: Ask which asset types are included, such as external networks, web applications, APIs, cloud resources and internal networks, and which need an add-on.
  • Scope changes: Ask how a newly found asset joins the scope, who confirms ownership, and whether it changes the fee.
  • Who tests: Ask which tests people perform, which agents perform, and what qualifications the human testers hold.
  • Confirmation and review: Ask how a finding is confirmed and to see a sample finding with its proof and retest history.
  • Retest terms: Ask whether retests are included and how a failed retest is handled.
  • Integrations: Ask which ticket systems are supported and whether status syncs both ways.
  • Audit outputs: Ask for a sample period report and whether an attestation letter or a separate independent test is available.
  • Data handling: Ask where test data and proofs are stored, which region they stay in, how long they are kept and how they are destroyed.
  • Exit: Ask whether the scope, findings and retest records can be exported in full when the term ends.

Public buyers often have a vetted route. In the US, GSA’s Highly Adaptive Cybersecurity Services buyer’s guide lists penetration testing as a subgroup and states that vendors listed in each subgroup have passed a technical evaluation for it. In the UK, the NCSC’s CHECK scheme serves government, the public sector and critical national infrastructure.

The split between people and agents decides what a subscription can find, and our guide to AI penetration testing explains where agents are strong and where a human tester leads.

A short trial against one application or domain shows more than a feature list. It reveals the quality of the proof, the noise in the findings and how the integration behaves with real tickets.

PTaaS Pricing Models

PTaaS is usually priced per term, with the unit set by the scope or by the testing effort. The models below are common, and many providers combine two of them.

Pricing model Unit Suits Points to check
Per asset or application Each domain, host group or application in scope for the term Estates with a stable count of systems How new assets found mid-term are charged.
Tiered subscription A band of assets or a package of features Organizations that want a predictable annual cost What moves the account to the next tier.
Credits or tester days A pool drawn down for human-led tests and deep dives Programs that mix continuous testing with periodic depth Whether unused credits expire and whether retests use them.
Platform fee plus tests A platform charge with each test or engagement priced on top Buyers moving from project testing Whether retests and triggered tests count as new tests.

Several factors move the cost under any model. They include the number and type of assets, authenticated testing with several user roles, the amount of human-led depth, internal network access, the reporting and attestation the subscriber needs, and the retest policy.

Quotes compare best on one written scope, with the same asset list and reporting needs given to every provider.

Reporting for Audits and Compliance

An auditor usually asks for a dated statement of what was tested and found, which PTaaS platforms typically produce as a period report generated from the record.

PCI DSS shows what such a report has to support. Requirement 11.4.1 calls for a defined penetration testing methodology that includes the following.

  • Accepted approaches: The methodology follows industry-accepted penetration testing approaches.
  • Coverage: Testing covers the entire cardholder data environment perimeter and critical systems.
  • Position: Testing runs from both inside and outside the network.
  • Segmentation: Testing validates any segmentation and scope-reduction controls.
  • Layers: Application-layer and network-layer penetration tests are both included.
  • Recent threats: The methodology reviews threats and vulnerabilities experienced in the last 12 months.
  • Risk handling: A documented approach assesses and addresses the risk posed by exploitable findings.
  • Retention: Penetration testing results and remediation results are retained for at least 12 months.

A subscription limited to internet-facing systems covers part of this list. The inside-the-network and segmentation elements typically need an internal test, which the provider or a second party may supply.

Independence is the other question an assessor asks. NIST SP 800-53 enhancement CA-8(1) asks for impartial testers, with no actual or perceived conflict of interest with the people who build, run or manage the tested systems, and PCI DSS requires organizational independence of the tester. A subscription delivered by the same firm that runs the systems may need a separate tester for the assurance part.

The PCI Security Standards Council’s Penetration Testing Guidance suggests that a retest report carry the date of the original test, the date of the retest, the original findings and the results of the retest, all of which a PTaaS record holds per finding. A useful period report adds the scope as it stood on the report date, the method, the tests run and their triggers, accepted risks, and what was left untested. What standards expect on penetration testing frequency is covered in its own guide.

Evidence From a Subscription

Each part of a PTaaS term leaves a record, and those records answer an auditor, a board or an insurer. The table lists the main ones.

Subscription record What it establishes Held in
Signed authorization and test policy Who approved testing, of which systems, under which limits The contract annex or the platform’s scope page, with signature dates
Scope history When each asset joined the scope and who confirmed ownership The platform’s asset list and its change log
Test log Each test with its trigger, time, source addresses and outcome The platform’s test history, exported per period
Finding record with proof The weakness, the affected asset and the evidence that it was exploited The platform finding page, linked from the ticket
Review note That a person checked a significant finding before release The finding’s history in the platform
Ticket history Who owned the fix and when status changed The owning team’s ticket system
Retest record The repeated proof and its result, with dates The finding’s history, mirrored in the ticket
Detection check Whether monitoring recorded the test traffic and raised an alert The SIEM or SOC case for the test window
Risk acceptance An owner’s decision to leave a finding open, with a review date The risk register, cross-referenced from the finding
Period report A dated summary of scope, tests, findings and retests Generated from the platform for the audit period

Records held only by the provider are a risk at the end of the term. The export terms agreed when choosing a provider decide whether this evidence stays with the organization.

PTaaS in a CTEM Program

A PTaaS subscription typically supplies the validation stage of a continuous threat exposure management program, and part of mobilization through retests. The CTEM guide explains the stages themselves, so this section covers only how a subscription plugs into them.

  • Shared scope: The CTEM scope and the subscription scope are kept as one list, so a scope agreed with a business owner is the scope the provider tests.
  • Priority list in: The program sends its ordered list of exposures to the provider, which tests the top of it ahead of the rolling schedule.
  • Results out: Each tested exposure returns as confirmed with proof, blocked by a control, or unreachable, and the program reorders its queue from those results.
  • Closure by retest: Mobilization closes an exposure only on a passed retest from the subscription.
  • Discovery feed: New assets the provider finds are passed back to the program’s inventory, so discovery and validation stay in step.

A Quarter of PTaaS at Acme-corp

acme-corp, a mid-sized online retailer, takes a subscription that combines AI agents for continuous tests with human testers for depth.

  1. Week 1, agreement. The CIO signs the authorization for acmecorp.com and its subdomains, the customer portal, its API and the remote access gateway. The checkout page, hosted by the payment processor, is excluded, and portal testing runs in agreed hours at a capped request rate.
  2. Week 1, integration. Findings for the portal and API route to the development team’s tracker, and findings for infrastructure route to the IT operations queue.
  3. Weeks 2 and 3, baseline. The baseline confirms a forgotten staging host that serves a configuration backup containing a database connection string, rated high. It also reports verbose error messages from the API that reveal internal paths, rated medium, and outdated TLS settings on a legacy mail gateway, rated low.
  4. Week 3, first fixes. No one claims the staging host, so the IT manager takes interim ownership and removes it from the internet. The retest the next morning finds the host unreachable and closes the finding, and the database password is rotated.
  5. Week 5, a new host. A certificate appears for promo.acmecorp.com, a campaign site built by an agency. After marketing confirms ownership, a first test finds that the site’s content management system allows anyone to register an account, rated medium, and the agency disables registration before the retest passes.
  6. Week 7, a new advisory. A vulnerability in the remote access gateway product is added to CISA’s Known Exploited Vulnerabilities catalog. The provider confirms the gateway runs an affected build, the critical finding goes to the IT manager by phone, and the patch is applied that day. The retest confirms the fixed build, and the SOC searches gateway logs for earlier exploitation attempts.
  7. Week 9, a release. A refund feature reaches the portal, and the subscriber draws on its tester days for a human-led test of the workflow. The tester finds that two parallel refund requests for one order both succeed, rated high, and the retest passes after the developers add a lock.
  8. Week 12, the audit report. The period report goes to the payment card assessor with the scope as it stood on the report date, every test and finding, retest dates and results, and the accepted risk on the mail gateway, which is due for replacement next quarter.

Every finding ended the quarter with a passed retest or an owner’s acceptance. A person found the refund flaw, and the agents tested the gateway and campaign site soon after each change.

Setting Up a Subscription

The first weeks set up the scope, people and workflow that every later test depends on.

  1. List the scope. Whoever leads security, often the IT manager in a smaller organization, records every domain, address range and application with its owner.
  2. Agree the test policy. System owners set exclusions, permitted hours, rate limits and the tests that need approval.
  3. Sign the authorization. An executive with authority over the systems signs, and any hosting provider’s testing rules are checked.
  4. Connect the ticket systems. The provider’s integration is set up for each owning team, with the rule that only a passed retest closes a ticket.
  5. Brief the SOC. The monitoring team receives the test source addresses and a contact at the provider.
  6. Work the baseline. Findings from the first pass are reviewed with owners in a weekly session until the backlog is under control.
  7. Set the audit outputs. The security lead confirms with the auditor or assessor what the period report must show and when it is due.

Pentest as a Service With ThreatDefence

A ThreatDefence subscription starts from an organization’s domains and a signed authorization, with no log source to connect first. The agents are part of the ThreatDefence AI SOC platform, which automates security monitoring, compliance and cyber risk management with a 24x7 SOC behind it.

  • Testing: The ASM agent keeps an inventory of the organization’s internet-facing domains, hosts, services and certificates, and the testing agent tests them on change, on schedule and on request.
  • Findings: Each finding lists the affected asset, its proof, fix steps and a severity rated by exploitability, and a SOC analyst checks the significant ones before release.
  • Retests: A reported fix sends the testing agent back to repeat its proof, and each problem keeps one tracked action item until it closes.
  • Earlier use: Where ThreatDefence also monitors the organization, the Defender and Hunter agents look through its telemetry for any earlier use of a confirmed weakness.
  • Partners: MSPs and MSSPs can deliver the subscription under their own brand, with a separate scope per client.

Higher-impact tests wait for approval. Internal networks, social engineering and attestation letters stay with an independent human-led test alongside. Agentic penetration testing explains the agents.

Frequently Asked Questions

What is penetration testing as a service (PTaaS)?

Penetration testing as a service, or PTaaS, is penetration testing bought as a subscription. One scope is agreed for the term, tests run on change, on a schedule and on request, and each finding is released through a platform as soon as it is confirmed. Fixes are retested inside the same agreement, and the platform keeps one running record of scope, tests, findings and retests.

How is PTaaS different from a one-off penetration test?

A one-off test examines named systems during a booked window and ends with a single report. PTaaS keeps the scope open for the term, tests new and changed systems as they appear, releases findings one at a time and retests when a fix is reported. Many organizations keep a periodic independent test for depth and use the subscription for the months between.

Does PTaaS meet a requirement for an independent penetration test?

Yes, in some cases, where the provider is organizationally independent of the systems and its method and records meet the standard’s wording. PCI DSS and NIST SP 800-53 both expect a tester free from conflicts of interest. Where a requirement names an attestation letter or internal network testing, a separate human-led test may still be needed, and the assessor decides.

What do PTaaS providers include in a subscription?

Most PTaaS providers include a baseline test, ongoing tests during the term, a platform showing findings with proof, retests of reported fixes and some form of period reporting. Coverage varies more than the headline suggests. Internal networks, authenticated testing with several roles, human-led depth and ticket integrations may be included, offered as add-ons or left out.

How are PTaaS subscriptions priced?

PTaaS is typically priced per term, using a count of assets or applications, a tier, a pool of credits or tester days, or a platform fee with tests priced on top. Cost rises with the number and type of assets, authenticated testing, human-led depth, internal access and reporting needs. Comparing quotes on one written scope shows what each one leaves out.

Are retests included in a PTaaS subscription?

Yes, in most subscriptions, though the terms vary. Some providers retest without limit, some count retests against an allowance, and some charge for retests after a set period from the original finding. Buyers should also confirm that a retest repeats the original proof and checks close variations, since a fix often covers only the reported path.

Can PTaaS findings be sent to an existing ticket system?

Yes, in most cases. Many PTaaS platforms integrate with common IT service management and developer tracking tools, or offer an API and webhooks. A good integration opens one ticket per problem, maps severity and owner to the ticket’s fields, keeps status in step both ways and starts a retest when the ticket is marked fixed. The proof itself is best kept in the platform behind a link.

Do PTaaS providers use human testers or AI agents?

Both, in most cases, in different proportions. AI agents and automation typically handle discovery, first tests of new hosts, advisory checks and retests, while human testers take internal networks, social engineering, chained attacks and business logic. People also review significant findings. Buyers should ask for a written list of which tests each performs.

Can a subscriber request a test ahead of a release?

Yes. Most subscriptions let the subscriber name a system for testing, and the request runs inside the agreed scope and test policy ahead of the rolling schedule. Notice helps, since the provider can plan human-led depth for a new workflow. Findings from a requested test arrive the same way as any other, each one as it is confirmed.

Talk to Our Experts

See how ThreatDefence automates security monitoring, compliance and cyber risk management in one platform.

    I'm an MSP, MSSP, or reseller

    By submitting this form, you accept the terms in our privacy policy.