What is Throughline?
Throughline is a CPA audit firm for AI-native companies, headquartered in Sydney with a US-registered CPA entity that issues its SOC 2 reports. Paul Wenham, who founded AssuranceLab, and Rob McAdam, who founded Pure Hacking and Sekuro, launched it in July 2026.
Its website is throughlineaudit.com. The firm also appears as Throughline Audit, Throughline CPA and Throughline LLP across its site and LinkedIn page, so ask which legal entity will sign your report.
It audits companies with AI in the product, in operations or in the compliance program, and still audits cloud-native companies with no AI in the product. It says its strongest fit is 1 to 1,000 people, mainly in Australia, New Zealand, the US, the UK and Europe, served by a team in Australia and the US.
Throughline audits only. It does not write policies or implement controls, which it presents as the basis of its independence, and readiness review happens inside the audit rather than as a separate paid project. The firm calls its method a calibrated audit: compliance goals, service commitments, company stage and technical scope decide which controls apply.
Is Throughline a legitimate SOC 2 auditor?
Throughline is a US-registered CPA firm, registered in Montana according to the firm, and it issues SOC 2 reports under AICPA attestation standards. Because it launched in July 2026, its first AICPA peer review is still pending and is due by January 2028.
That review will cover Throughline’s first 12 months of work, so it cannot finish sooner. A pending first review is normal for a firm this new, but there is no result to check until the AICPA’s public file shows one. Our guide to AICPA peer review and audit quality explains what a result tells you.
Throughline does not publish the names of its signing CPAs; it introduces them during the sales process. Ask for their names, their licence states and the firm’s licence number, and check them with the state boards before you sign.
The track record behind Throughline belongs to its founders’ earlier work. Paul Wenham managed SOC reporting engagements at PwC, the firm says, and then built AssuranceLab, a peer-reviewed CPA firm that supported more than 1,200 technology companies before Sensiba acquired it. Those figures describe AssuranceLab, not Throughline. Throughline itself has no published client reviews or named clients yet; reviews of AssuranceLab’s work are listed under the Sensiba name.
What makes a Throughline SOC 2 audit different for AI companies?
Throughline scopes and tests a SOC 2 around where AI sits in a client’s product and operations. Its audits cover six AI control areas within the standard Trust Services Criteria: agent access, AI-enabled changes, monitoring, governance, data handling and incident response.
| Control area | What Throughline examines |
|---|
| Agent access | Permissions held by non-human identities, approved, reviewed and removed with the same care as staff access |
| AI-enabled changes | A record of what an agent proposes or changes, with testing, approval and deployment safeguards scaled to the possible impact |
| Monitoring | Agent activity, failures and unusual behaviour, plus changes to models, prompts, tools and connected workflows |
| Governance | Ownership, permitted uses, risk review, human oversight and escalation for each AI use |
| Data handling | What enters each model or workflow, where it goes, how long it stays, and which sensitive data is restricted |
| Incident response | AI-related security and service events inside detection, containment and investigation |
The firm separates AI in the product from AI in operations. Product AI that processes customer data, generates outputs or acts in a customer’s environment can change the report’s scope, system description and commitments. Operational AI, such as coding agents and internal assistants, sits behind the product but can still reach in-scope systems, so it needs proportionate controls too.
The report has limits. In Throughline’s words, “SOC 2 does not certify an AI model.” It gives customers independent evidence about the controls around the system that builds, runs and monitors an AI service, not about whether a model is accurate, unbiased or safe.
Throughline also uses AI for its own documentation, evidence analysis and audit administration. For control detail, see our guides to SOC 2 for AI companies and SOC 2 controls for AI agents and MCP, or the firm’s SOC 2 for AI-native companies page.
How much does Throughline charge for SOC 2?
Throughline charges one annual SOC 2 fee that covers both Type 1 and Type 2, starting at US$7,500 a year for companies of 1 to 10 people. The fee steps up with company size; above 500 people, the firm quotes each engagement.
The starting price is firm-confirmed as of October 2026. For the fee at your size, request a quote; the quote confirms headcount, scope and report structure. Each package includes the Security, Availability and Confidentiality criteria, readiness review, a dedicated senior practitioner for the year and the signing-CPA review.
Because one fee buys both reports, this page shows one package price rather than separate Type 1 and Type 2 prices. A standalone Type 1 costs half the package fee, and Throughline credits it in full toward the package if you go on to a Type 2 in the same year.
Privacy and Processing Integrity cost extra. HIPAA, GDPR and SOC 1 reports are priced separately, with a discount when you buy one together with SOC 2. Extra products, a second report or a carved-in subservice organisation can add to the fee, which the firm confirms after scoping.
The annual fee is billed in 12 equal monthly payments, in US dollars, or in Australian dollars and euros for clients in Australia and Europe. Clients can cancel at any time. If no report has been issued in the contract year, no remaining fees are payable; once a report has been issued, the remaining monthly payments for that contract year stay payable.
Does Throughline certify ISO 42001 and ISO 27001?
Yes. Throughline also runs ISO 27001 and ISO 42001 certification audits and issues the certificates itself, as its own certification body. That body is not yet accredited: Throughline is pursuing accreditation from the International Accreditation Service (IAS) and targets roughly January to July 2028, 18 to 24 months after its July 2026 launch.
Accreditation is an independent check of the certification body itself, against ISO/IEC 17021-1 plus the standard for each scheme: ISO/IEC 27006-1 for ISO 27001 and ISO/IEC 42006 for ISO 42001. For a first accreditation, the accreditation body observes the certification body’s Stage 1 and Stage 2 audits of at least one client, so a new certification body has to start auditing clients before it can be accredited. AssuranceLab, Paul Wenham’s earlier firm, was an IAS-accredited certification body, according to Throughline.
Initial certification runs as a Stage 1 and Stage 2 audit, followed by annual surveillance. An ISO 42001 audit starts from an inventory of the AI a company builds, buys and embeds. The certificate covers the AI management system; Throughline’s terms state that it does not certify any specific AI system, model, output or use case.
A customer can check an accredited certificate in the accreditation body’s public directory, and accredited bodies must also upload their certificates to IAF CertSearch. ISO’s certification page says accreditation is not compulsory and that non-accreditation does not necessarily mean the certification body is not reputable, but if a customer, tender or regulator requires an accredited certificate, Throughline cannot meet that requirement yet. The firm says it raises this at the start of an engagement.
Ask your largest customers what they accept before you commit. Ask Throughline which legal entity will issue your certificate, and whether it would be reissued under accreditation once that is granted. Our guides to ISO 27001 certifying bodies and ISO 42001 certification cost explain how to check accreditation and what certification typically costs.
Who leads Throughline?
Throughline is led by its two co-founders: chief executive Rob McAdam, who founded Pure Hacking, CXO Security and Sekuro, and Paul Wenham, a former Big 4 audit manager who founded AssuranceLab. A named senior practitioner leads each client engagement.
Paul Wenham (co-founder) is a former Big 4 audit manager; Throughline says he managed SOC reporting engagements at PwC. He founded AssuranceLab, a SOC 2 and ISO audit practice that Throughline says supported more than 1,200 technology companies. He became a Sensiba partner after Sensiba acquired it in April 2025, and left Sensiba in March 2026.
Rob McAdam (co-founder and CEO) founded Pure Hacking, CXO Security and Sekuro, and spent two decades building offensive-security and cyber consulting practices for enterprise and government clients.
Throughline says its practitioners average more than 10 years of experience, with no junior hand-offs and no offshore delivery.
How is Throughline different from AssuranceLab and Sensiba?
Throughline is a separate, new firm. Paul Wenham founded AssuranceLab, which Sensiba LLP acquired in April 2025, then launched Throughline with Rob McAdam in July 2026. Throughline did not inherit AssuranceLab’s peer review, accreditation or client base.
AssuranceLab’s practice now sits inside Sensiba, a San Jose CPA firm, where AssuranceLab co-founder Nicholas Lew Ton is a partner. Throughline carries over its founder’s experience and methods, not credentials a buyer can look up.
The difference for a buyer is the evidence available today. Sensiba has a public AICPA peer-review Pass and ANAB accreditation for ISO 27001 and ISO 42001 certification; Throughline has neither yet. If a customer contract requires either before 2028, Sensiba can meet it now. If you want a small, senior-led team focused on AI-native companies and can accept pending credentials, Throughline is the newer firm from one of AssuranceLab’s founders.
How long does a Throughline SOC 2 audit take?
Throughline tells clients to plan for three to six months from start to report, mostly client-side preparation, with the audit itself typically taking four to six weeks. Some engagements have finished in one to two weeks; the firm does not promise that pace.
A Type 2 also needs its observation period, usually 3 to 12 months. Throughline often recommends a Type 1 first, which gives customers a formal report sooner and leaves room for a six- or twelve-month Type 2 period rather than a rushed three months.
Throughline describes the engagement as five stages:
- Scoping of the system boundary and criteria. Cloud providers such as AWS are normally carved out.
- Readiness inside the audit. Issues found before a Type 1 can usually be fixed and retested before its as-of date.
- Pre-audit analysis, an automated review of the controls with feedback before testing.
- Audit window. Type 2 testing can be split across the period, with control issues reported as found.
- Report and signing-CPA review, planned around one week once information is complete, not a guaranteed issue date.
During a Type 2, Throughline can issue a confirmation-of-engagement letter and factual status letters, gives clients a bridge letter template to send their customers between report periods, and will join customer calls to explain the report. None of this guarantees that a customer will accept a report, an exception or a Type 1.
Throughline is platform-agnostic. It is a Drata partner, works inside a client’s Drata or Vanta account, and can run the audit with no GRC platform, taking evidence from Confluence, Jira, ClickUp, Google Drive, GitHub and similar systems. Clients do not need to buy or switch software.
Where no export exists, the firm says it collects evidence through browser-agent workflows and its own practitioners, and adapts controls to the client’s architecture instead of a fixed control list.
Who is Throughline a good fit for, and when is it not?
Throughline fits AI-native and cloud-native companies of roughly 1 to 1,000 people that want SOC 2 testing built around their agents, models and existing tools. It fits less well when a customer needs a completed peer review now.
Shortlist Throughline if:
- AI is in your product or agents act in your environment, and enterprise risk teams ask how you control them.
- You are in Australia or New Zealand and want an AICPA SOC 2 from a US-registered CPA firm, with a Sydney-headquartered team in your time zone. Compare other SOC 2 auditors in Australia.
- You want readiness inside the audit, one package for Type 1 and Type 2, and no GRC platform requirement.
Look elsewhere if:
- A customer, investor or tender requires a completed peer review before January 2028, or an accredited ISO certificate before Throughline’s accreditation is granted.
- Your procurement team wants named client references or published reviews.
- You need policies written or controls built. Throughline only audits, so pair it with a readiness firm.
- You need HITRUST, PCI DSS, FedRAMP or a penetration test, none of which Throughline lists.
- You want an audit that asks little of your team. The firm names buyers seeking a rubber-stamp audit as its one poor fit.
The closest comparison is Chiaro, a two-person Austin CPA firm for AI-native companies that also launched in 2026 and is also awaiting its first peer review. Chiaro fits teams of about 1 to 20 people, lists lower firm-confirmed Type 2 fees and runs the examination on its own MCP platform. Throughline suits larger teams, teams that want the audit run in their existing tools rather than on a separate platform, and teams that need ISO or GDPR work alongside SOC 2. See all SOC 2 auditors for AI companies.