Table of Contents

Reach SOC 2 Compliance in 6 Weeks or Less.

  /

  / VAPT Report Guide: What It Includes & How to Write One

VAPT Report Guide: What It Includes & How to Write One

A Vulnerability Assessment and Penetration Testing report is the final deliverable where weeks of security testing either turn into action or quietly fade away in a company’s digital archive. The testing finds the holes, and the report decides whether anyone fixes them.

Get it wrong, and you have an expensive PDF that satisfies an auditor and protects nobody. Get it right, and you have a prioritised plan that tells your team exactly what to fix first and why it matters, saving you a lot of money in avoided security breaches in the long run.

This guide covers what a VAPT report is, what belongs in it, how to write one that holds up under scrutiny, and how it ties into the certifications most businesses actually care about.

VAPT Report

What Is a VAPT Report?

VAPT stands for Vulnerability Assessment and Penetration Testing. The report is the document that captures everything the testing uncovered: the weaknesses, how serious each one is, which an attacker could realistically exploit, and what to do about them.

The two halves do different jobs.

A vulnerability assessment is broad and largely automated. It scans systems, networks, and applications to produce a prioritised list of known weaknesses, without trying to exploit them.

Penetration testing is narrow and manual. A skilled tester takes selected weaknesses and tries to exploit them, chaining flaws together the way a real attacker would, to prove what damage is actually possible. One gives you visibility. The other gives you validation. A strong VAPT report fuses both into a single picture of real risk rather than theoretical exposure.

 Vulnerability AssessmentPenetration Testing
ApproachBroad, mostly automated scanningFocused, manual exploitation
GoalIdentify known weaknesses at scaleValidate real-world impact
OutputPrioritised list of weaknessesExploited findings with proof of concept
AnswersWhat might be wrong?What can an attacker actually do?

Reach SOC 2 Compliance in 6 Weeks or Less

Schedule Your Free SOC 2 Assessment Today

What Is the Objective of a VAPT Report?

The objective is not to list vulnerabilities. Any scanner can produce a list. The objective is to turn raw findings into decisions: what to fix, in what order, and how much each issue matters to the business.

A good report does three things at once. It gives executives a clear read on risk and the cost of ignoring it. It gives engineers the technical detail and reproduction steps they need to fix each issue. And it creates a point-in-time record proving that testing happened, which auditors, regulators, and customers all ask to see. The same document has to serve a boardroom and a bug queue, which is exactly why structure and audience awareness matter so much.

 

Who Needs a VAPT Report?

Almost any organisation that runs internet-facing systems or handles sensitive data benefits from one. Three groups need it most.

Organizations Pursuing or Maintaining Compliance

This is the most common trigger. Frameworks such as PCI DSS, SOC 2, ISO 27001, and GDPR all expect some form of security testing, and a VAPT report is the cleanest way to evidence it. For regulated businesses, the report is not optional documentation. It is the artefact an assessor reviews to decide whether a control is actually working, and a missing or stale report can stall an entire certification.

Organizations of Any Size

Size offers no protection. Automated attacks scan the entire internet indiscriminately, and a small company with an exposed admin panel is a softer target than a large enterprise with a mature security team. Regular testing matters most after meaningful change: a new product launch, a cloud migration, an acquisition, or rapid headcount growth. Each of those expands the attack surface faster than most teams update their defences.

Clients and Business Partners

Increasingly, the report is a sales document. Enterprise buyers send security questionnaires before they sign, and “do you conduct penetration testing, and can we see a summary?” is now a standard line item. A clean, customer-facing summary of a VAPT report shortens sales cycles and builds trust. Its absence becomes a gap that procurement teams probe directly.

Worth Knowing: Enterprise Vendor Assessments

Enterprise vendor assessments such as SIG and CAIQ routinely ask about penetration testing frequency, findings, and remediation. A polished report you can share on request often does more for a deal than another case study, because it answers a security reviewer's question before they have to chase you for it.

The Anatomy of a VAPT Report: Key Elements

Formats vary by tester and by standard, but credible reports share the same seven building blocks.

  1. Executive Summary. A non-technical overview for leadership. It states the overall risk posture, the headline findings, and the business impact in plain language. For many executives this is the only section they will read, so it has to stand on its own.
  2. Methodology, Scope, and Tools Used. What was tested, what was deliberately excluded, which standards were followed (commonly OWASP, PTES, or NIST Special Publication 800-115), which tools were used, and the dates of the engagement. Scope is what defines the boundary of every claim the report can make.
  3. Scan Results and Details of Tests Performed. The summarised output of automated scanning alongside the specific manual tests carried out, giving reviewers a clear view of coverage.
  4. Detailed Findings and Vulnerabilities. The core of the document. Each finding gets a description, the affected asset, a severity rating, supporting evidence, and clear reproduction steps so the fix can be verified later.
  5. Risk Assessment Profile. Each vulnerability rated by severity, exploitability, and business impact, most often scored with a framework such as the Common Vulnerability Scoring System. This is what lets a team prioritise rationally instead of fixing whatever looks scariest.
  6. Remediation Planning and Recommendations. Specific, prioritised, actionable fixes, ideally with suggested timelines and owners. Vague advice like “harden the server” fails here. “Disable TLS 1.0 on these three endpoints” succeeds.
  7. Appendices and Supporting Evidence. Screenshots, request and response captures, payloads, proof-of-concept artefacts, and raw scanner output. This is the material that turns assertions into proof.

Pro Tip: Writing the Executive Summary

Write the executive summary last, and write it for a reader who will never get past it. If your CEO read only that one page, would they understand the single most important risk and what you need from them to close it? If not, the summary is not finished.

VAPT Report - 5 Steps

How to Write a VAPT Report

Step 1: Understand Your Purpose and Audience.
A report written for an internal engineering team looks nothing like one written for a board or an auditor. Decide upfront whether the goal is a point-in-time health check, compliance evidence, or due diligence, then set the depth and language accordingly.

Step 2: Gather Necessary Information.
Pull together the defined scope, the asset inventory, scan output, exploitation logs, and every piece of evidence captured during testing. Collect it as you go, not at the end, when details are already fading.

Step 3: Structure the Report.
Follow a logical flow: executive summary, then methodology and scope, then detailed findings, then the risk profile, then remediation, then appendices. A reader should be able to move from “how bad is it” to “what do I fix first” without hunting.

Step 4: Attach the Necessary Evidence and Proof of Concept.
Each significant finding needs evidence: a screenshot, a captured request, or reproduction steps. Evidence is what separates a credible report from an unverifiable claim, and it is the first thing a sceptical reviewer looks for.

Step 5: Final Review.
Check severity ratings for consistency, strip out false positives, confirm every recommendation is actionable, and make sure the language fits the audience. A single inflated or unverifiable finding undermines confidence in the whole report.

Insider Note: Auditors and PCI QSAs are now trained to spot an automated scanner export dressed up as a penetration test. A report with no manual analysis, no chained findings, and no evidence the tester went beyond running a tool is increasingly rejected as insufficient. The narrative of how findings connect is often what proves a human did the work.

Reach SOC 2 Compliance in 6 Weeks or Less

Schedule Your Free SOC 2 Assessment Today

VAPT Report Example: A Brief Walkthrough

Imagine a tester finds a staging admin login that is publicly reachable and has no multi-factor authentication. On its own, a scanner would flag it as a medium-severity exposure. The penetration tester goes further: they discover the staging credentials are reused in production, log in, and reach a database holding live customer records.

That single chain shows up differently in each part of the report.

  • The executive summary states it in one line: an exposed login allowed access to live customer data, the highest-priority risk found.
  • The detailed finding documents the endpoint, the missing MFA, the credential reuse, and the exact steps to reproduce it.
  • The risk profile rates it critical, because exploitability is high and the business impact is a potential data breach.
  • The remediation section says precisely what to do: remove public access to staging, enforce MFA, and rotate the shared credentials.
  • The appendix holds the screenshots and request logs that prove every step. One finding, six sections, no ambiguity about what to fix first.

 

VAPT Report Templates and Automation Tools

Downloadable VAPT Report Templates

You do not have to design a report from scratch. Established methodologies publish free structures you can adapt: the OWASP testing guidance, the Penetration Testing Execution Standard (PTES), and NIST Special Publication 800-115 all describe how to organise scope, findings, and reporting. Starting from a recognised structure also signals rigour to auditors, who tend to trust reports that map to a known standard.

Reporting Automation and Management Tools

Tooling generally falls into three categories: vulnerability management platforms that aggregate and track findings over time, pentest reporting tools that turn evidence into structured documents, and ticketing integrations that push findings straight into systems like Jira so remediation gets owned and tracked. Automation speeds up the mechanical parts of reporting, but it does not replace the analyst’s judgement on severity and business context. The fastest way to produce a useless report is to let a tool decide what matters.

 

VAPT Reports and Compliance

This is where a VAPT report stops being a security artefact and becomes a business one. Different frameworks treat penetration testing differently, and getting the distinction right matters. Claiming a framework “requires” a penetration test when it does not is a common, avoidable mistake that erodes credibility with informed buyers and auditors alike.

Where VAPT Is Explicitly Required (PCI DSS)

PCI DSS is the clearest mandate of any major standard. Under the current version maintained by the PCI Security Standards Council, Requirement 11.4 explicitly calls for both internal and external penetration testing at least every twelve months and after any significant change, supported by a documented methodology and segmentation testing. If you store, process, or transmit cardholder data, a missing or out-of-date test is a straightforward audit failure, not a matter of interpretation.

Important: Penetration testing moved from Requirement 11.3 in PCI DSS v3.2.1 to Requirement 11.4 in v4.0, and the future-dated requirements became mandatory on 31 March 2025. Any template, checklist, or internal note still citing “11.3” for penetration testing is out of date. Reviewers notice, and it signals the rest of the document may be stale too.

Where VAPT Is Expected as Evidence (SOC 2, ISO 27001)

Neither framework names penetration testing as a hard rule, which surprises a lot of teams. The AICPA’s Trust Services Criteria that underpin SOC 2 never use the phrase “penetration test.” In practice, though, most auditors treat it as expected evidence for criteria such as CC4.1 (monitoring activities) and CC7.1 (vulnerability identification), and showing up to a Type II audit without one usually invites observations or requests for more evidence. The working standard is at least one test a year, falling inside the audit observation period.

ISO 27001 takes a similar shape. The standard does not mandate a pen test, but the technical vulnerability management control (Annex A 8.8 in the 2022 revision) and the core risk-assessment requirements make penetration testing the most natural way to demonstrate that those controls operate. In both cases the report is the evidence that turns a written policy into something an auditor can verify.

Where VAPT Supports Compliance (GDPR and others)

GDPR sits at the softest end of the spectrum. Article 32 requires a process for regularly testing, assessing, and evaluating the effectiveness of security measures, without ever naming a technique. Regulators and data protection authorities consistently treat penetration testing as a strong way to satisfy that obligation, especially for organisations processing sensitive data at scale. The report becomes part of the evidence trail that a breach response will be judged against.

Framework

Is VAPT required?

Where the expectation lives

Typical frequency

PCI DSS

Yes, explicitly

Requirement 11.4

Annual + after major change

SOC 2

No, but expected

Criteria CC4.1, CC7.1 (auditor evidence)

Annual, within audit period

ISO 27001

No, but expected

Annex A 8.8 + risk assessment

Annual + after major change

GDPR

No, but supports it

Article 32 (regular testing)

Risk-based, commonly annual

Using Your VAPT Report in a SOC 2 or ISO 27001 Audit

To make the report carry weight in an audit, map each finding to the specific control or criterion it touches, keep the test inside the audit period, and include retest evidence showing critical findings were fixed. Independent third-party testing is almost always stronger evidence than internal testing, because it demonstrates objectivity. If you are working toward either certification, it is worth aligning your testing scope with your audit scope from the start. Axipro covers this in more detail on our SOC 2 compliance and ISO 27001 certification pages.

Benefits of a VAPT Report

Protection from Cyber Threats. The direct benefit. A report surfaces the weaknesses an attacker would find first, giving you the chance to close them before anyone else does.

Compliance with Security Standards. A single well-run test and its report can serve as evidence across multiple frameworks at once, turning one engagement into PCI DSS, SOC 2, ISO 27001, and GDPR coverage.

Better Incident Management. The remediation workflow a report triggers (identify, assess, fix, verify) mirrors incident response. Teams that practise it on findings respond faster to real incidents.

Improved Market Perception. Being able to show a clean report, or a credible remediation track record, signals maturity to customers, partners, and investors. In competitive deals, that signal can be the difference.

 

A VAPT Report Doesn’t Reduce Risk. Remediation Does.

The Gap Between Finding and Fixing

A report names the problem. It does not solve it. The risk persists, in full, until the fix actually ships. Plenty of organisations accumulate a shelf of reports while the same critical findings reappear year after year, because producing the document and closing the issue are two completely different pieces of work.

Why Businesses Become Focused on Reports Over Outcomes

The report is the thing that gets paid for, shown to auditors, and attached to a deal. So it quietly becomes the goal. Compliance incentives reward having a report, not closing what it contains, and filing a PDF is far easier than coordinating engineering time to remediate. The result is a slow drift from security as an outcome to security as paperwork.

What Effective VAPT Looks Like Beyond the Report

Effective programmes treat the report as a starting point, not a finish line. Findings feed a tracked remediation workflow with named owners and deadlines. Critical issues get retested to confirm the fix worked. And the testing cadence moves away from a single annual snapshot toward something closer to continuous, because attackers do not wait twelve months between attempts. The report still matters. It just stops being the point.

Reach SOC 2 Compliance in 6 Weeks or Less

Schedule Your Free SOC 2 Assessment Today

Conclusion

A VAPT report is only as valuable as the action it drives. Done well, it translates technical findings into clear, prioritised decisions, evidences your compliance posture across PCI DSS, SOC 2, ISO 27001, and GDPR, and gives customers a reason to trust you. Done badly, it is a document that proves you looked without proving you fixed anything. The structure and standards matter, but the real test of a report is simple: did the risks it described actually get smaller?

Frequently Asked Questions

What is the main purpose of a VAPT report?

To turn the findings of security testing into prioritised, actionable decisions, while creating a point-in-time record that proves testing happened. It serves both technical teams who need to fix issues and leadership, auditors, and customers who need assurance.

At least annually for most frameworks, and after any significant change such as a major release, infrastructure shift, or acquisition. PCI DSS requires testing at least every twelve months, and SOC 2 and ISO 27001 auditors expect a similar cadence. Higher-risk environments increasingly move toward continuous testing.

Internally, the security and engineering teams who will remediate, plus the leadership who own the risk. Externally, auditors and assessors who need evidence, and, in summary form, enterprise customers who request it during procurement.

Yes, but usually as a controlled summary rather than the full technical document. The detailed findings and proof of concept are effectively a map of your weaknesses, so the complete report is shared under NDA or replaced with a sanitised attestation for customers and partners.

Clear prioritisation, evidence behind every finding, remediation advice specific enough to act on, and language tailored to its audience. Above all, it is effective only if the risks it identifies actually get fixed and retested.

Severity, exploitability, and business impact. Severity reflects how serious the flaw is, exploitability reflects how realistically an attacker could use it, and business impact reflects what it would cost the organisation. Frameworks such as CVSS combine these into a single score so findings can be ranked consistently.

A vulnerability assessment report is broad and lists known weaknesses identified by mostly automated scanning, without confirming whether they can be exploited. A penetration test report is narrower and documents weaknesses that a human tester actually exploited, with proof of concept and real-world impact. A VAPT report combines both: the breadth of the assessment and the validation of the test.

Neither framework strictly requires penetration testing in its written text. SOC 2’s Trust Services Criteria and ISO 27001’s controls never mandate it by name. In practice, auditors for both expect it as the most credible evidence for vulnerability management and control effectiveness, so for most organisations it is effectively necessary even though it is not technically mandatory. PCI DSS, by contrast, does require it explicitly.

Axipro Author

Picture of Pedro Dias

Pedro Dias

Pedro has been writing online for over 10 years. With experience in all things programming, cyber security, and compliance, he is our editor-in-chief at Axipro.

Blog Highlights

Explore More Articles

Legacy threat modeling frameworks such as STRIDE were designed for software that behaves the same way over and over again. Agentic AI does no such thing. It can rewrite its own plan mid-task, call external tools, negotiate with other agents, and produce a different output from identical input. MAESTRO exists because none of the legacy threat modeling frameworks were built to handle that. MAESTRO stands for Multi-Agent Environment, Security, Threat, Risk, and Outcome. It is a seven-layer threat modeling framework created specifically for agentic AI systems, and it has become the closest thing the industry has to a standard method for reasoning about agent security. Understanding MAESTRO in the Context of Agentic AI What MAESTRO Stands For Each word in the acronym carries meaning. Multi-Agent Environment signals that the framework models entire ecosystems of interacting agents, not a single model behind an API. Security, Threat, Risk covers the core discipline: identifying attack surfaces, cataloging threats, and assessing likelihood and impact. Outcome is the part most frameworks skip. MAESTRO asks what an attack actually produces in the real world, because an autonomous agent with tool access turns a compromised prompt into a compromised action. The Origin of MAESTRO (Cloud Security Alliance) The Cloud Security Alliance published MAESTRO in February 2025. Its creator is Ken Huang, Co-Chair of the CSA AI Safety Working Groups and CEO of DistributedApps.ai. The CSA has since applied the framework publicly to real systems, including OpenAI’s Responses API and Google’s A2A protocol, which gives practitioners worked examples rather than just theory. The framework is openly published, and the CSA maintains an official companion tool, the MAESTRO Threat Analyzer, on GitHub. SOC 2, ISO 27001 and HIPAA done for you. Fixed fee, 100% audit pass rate. Audit-ready in 6 weeks. Not 6 months. Schedule Free Assessment Why Traditional Frameworks Fall Short for Agentic AI STRIDE, PASTA, LINDDUN, and OCTAVE all share a founding assumption: the system under analysis follows predictable logic with clearly defined boundaries. You draw the data flow diagram, mark the trust boundaries, and enumerate threats against components that behave deterministically. Agentic AI breaks every part of that assumption. Unique Security Challenges of Autonomous Agents Agents introduce three properties that legacy models cannot express. Non-determinism means the same input can produce different behavior, so you cannot enumerate execution paths in advance. Autonomy means the agent makes decisions and takes actions without a human approving each step, which collapses the usual assumption that a person sits between intent and execution. And in multi-agent systems there is often no stable trust boundary: agents delegate to other agents, consume tool outputs from external servers via protocols like the Model Context Protocol (MCP), and update their own memory and goals at runtime. The Gap Between Legacy Frameworks and Agent-Based Systems The practical consequence is coverage gaps. STRIDE has no category for goal manipulation, where an attacker gradually steers what an agent is trying to achieve. PASTA assumes attacker objectives and data flows are fixed, which fails for systems that learn and adapt during operation. LINDDUN addresses privacy but says nothing about agent collusion or memory poisoning. A threat model built purely on these frameworks will pass review and still miss the attacks that matter most in an agentic deployment. How MAESTRO Addresses Agentic-Specific Risks MAESTRO does not discard the older frameworks. It extends them with a layered reference architecture, an AI-specific threat catalog for each layer, and, critically, explicit analysis of how threats propagate between layers. That cross-layer lens is the framework’s real contribution, because most serious agentic incidents are chains: poisoned data influences a model, the model misleads an agent, and the agent takes an unauthorized action three layers away from where the attack started. The Seven Layers of the MAESTRO Framework MAESTRO decomposes any agentic system into seven layers, each with its own threat landscape. Layer 1: Foundation Models The core LLMs or other models the agents reason with. Threats here include adversarial examples, model extraction, backdoored weights, and jailbreaks that bypass safety training. If the model is a third-party API, supply chain risk lives at this layer too. Layer 2: Data Operations Everything the agent ingests, stores, and retrieves: training data, RAG pipelines, vector databases, and agent memory. Data poisoning and memory tampering are the signature threats at this layer, and they are especially dangerous because a poisoned memory persists across sessions and keeps shaping future decisions long after the initial attack. Layer 3: Agent Frameworks The orchestration software that turns a model into an agent: LangChain, CrewAI, AutoGen, custom planners, and tool-calling logic. Threats include prompt injection through tool outputs, insecure tool definitions, and manipulation of the planning loop itself. Layer 4: Deployment Infrastructure The servers, containers, and cloud services the agents run on. The CSA’s threat catalog here reads like traditional cloud security with an agentic twist: compromised container images carrying malicious agent code, Kubernetes orchestration attacks, denial of service against agent runtimes, and tampering with Infrastructure-as-Code templates that provision agent resources. Layer 5: Evaluation and Observability The systems that monitor, evaluate, and debug agent behavior. This layer is often forgotten, and attackers know it. The CSA specifically flags poisoning observability data: manipulating the telemetry fed to monitoring systems so that incidents stay hidden from security teams while malicious activity continues. Layer 6: Security and Compliance MAESTRO treats this as a vertical layer that cuts across all others: identity and access management, guardrails, policy enforcement, and compliance controls. Threats include permission escalation, guardrail bypass, and compromise of the security agents themselves in architectures where AI enforces policy on other AI. Layer 7: Agent Ecosystem The environment where agents interact with users, other agents, and marketplaces. This is where the genuinely novel threats live: agent impersonation, misleading agent capability cards, tool squatting, and collusion between agents to achieve outcomes no single agent was authorized to pursue. Insider Note: In real assessments, Layers 5 and 6 expose the maturity gap fastest. Most teams’ shipping agents can describe their model and their orchestration framework in detail, then

EU AI Act Hiring Map

AXIPRO STUDY New Study: Europe is hiring AI builders faster than AI governance professionals Axipro analyzed 3,519 AI-related job postings across eight EU countries. For every professional hired to keep AI lawful, safe and accountable, nearly seven were hired to build more of it, and the gap is widest exactly where you’d least expect. Take EU AI ACT READINESS QUIZZ 16 AI Builders : 1 AI Governors Sweden — Europe’s widest AI governance gap 3,519 Job Postings Analyzed 8 EU Countries 2 Role Categories: Builders vs Governors July 2026 Date of Job Postings Analyzed The findings Finding 1: Sweden hires 16 AI builders for every 1 person to govern them Throughout our data-set we found the same pattern across all eight countries: the more a nation hires to build AI, the less it hires to govern it. France runs eleven builders to every governor. Even Ireland, the most balanced in Europe, looks responsible mainly because the US tech giants headquartered there import global-governance discipline under overlapping DORA and AI Act pressure.  3.5→16 builders hired per governor, Europe’s most balanced country to its least. Ireland 3.5 Germany 5.7 Spain 6.0 Italy 7.1 Netherlands 7.2 Belgium 7.9 France 11.4 Sweden 16:1 0 4 8 12 16 Builders hired per AI governor Source: Axipro, 2026 Sweden has one of the strongest engineering cultures in Europe. It also carries the widest governance gap we measured: sixteen AI builders hired for every person hired to govern them. France sits close behind at eleven to one. The most balanced country, Ireland at 3.5 to one, looks responsible for a reason that has little to do with virtue. The US tech giants headquartered in Dublin import global governance discipline, and they do it under the combined weight of the AI Act and DORA, the EU financial-sector resilience regime in force since January 2025. Engineering strength does nothing to close a governance gap, and it may widen it. A country that ships AI faster produces more systems that fall under the Act’s scope and, on this evidence, fewer people positioned to document, monitor, and defend them. Being good at building AI offers no protection against governing it badly. The countries most confident in their technical talent are running the largest deficit against the law. Explore AI governance hiring by country Click any country to see how many AI builders it hires for every governance professional, and where it ranks against the rest of Europe. Germany — 5.7 builders per governorDE France — 11.4 builders per governorFR Spain — 6.0 builders per governorES Italy — 7.1 builders per governorIT Netherlands — 7.2 builders per governorNL Belgium — 7.9 builders per governorBE Ireland — 3.5 builders per governorIE Sweden — 16 builders per governorSE 3.5 — balanced 16 — widest gap Source: Axipro, 2026 Sweden 16builders for every governance professional Rank 1 of 8 · 20 governance roles vs 319 builder roles posted Only 30% of the AI governance roles name the AI Act Share this Embed this map Copy & paste — links back to Axipro Copy embed code Branded, one paste, backlink included. × Share this country insight Share this AI governance gap X / Twitter LinkedIn Facebook WhatsApp Bluesky Email Copy link Choose a platform or copy the link. A view of the same country-level dataset behind the interactive map: governance roles, builder roles, builder-to-governance ratio, and the share of governance postings that name the EU AI Act. AI governance jobs Europe statistics by country: governance roles, builder roles, builder-to-governance ratio and AI Act mention percentage. Country Governance roles Builder roles Builder-to-governance ratio AI Act mention % Sweden 20 319 16.0:1 30.0% France 39 443 11.4:1 38.5% Belgium 38 299 7.9:1 39.5% Netherlands 61 439 7.2:1 31.1% Italy 40 284 7.1:1 45.0% Spain 64 384 6.0:1 28.1% Germany 88 501 5.7:1 27.3% Ireland 96 335 3.5:1 14.6% Source: Axipro analysis of AI builder, governance and compliance job postings across eight European countries. “AI Act mention %” is the share of governance postings that explicitly name the EU AI Act. We analyzed thousands of AI-related job postings across eight EU countries and split them into two camps: the people hired to build AI systems, and the people hired to govern them. The ratio between those two groups tells you how seriously a country, a sector, or a company is treating the law that now governs both. Three numbers stood out.  Finding 2: Europe is hiring for a law it won’t say out loud Europe spent years drafting the AI Act. It cleared the European Parliament, survived the Digital Omnibus revisions, and now carries penalties that reach €35 million or 7% of global turnover for the most serious breaches, a ceiling that makes GDPR fines look modest. Yet fewer than three in ten of the governance roles created to handle it actually name the law in the job description. Among builder roles, the figure collapses to one in twenty-five. More than 7 in 10 Governance job descriptions do not mention the EU AI Act. This number rises to 9 in 10 for all AI job descriptions. Despite hiring for governance, risk, privacy, and compliance roles, most employers are not yet translating the EU AI Act into explicit job requirements. That disconnect should stop you. The people being hired to make Europe compliant are, for the most part, not being hired against the Act by name. They are titled around adjacent ideas: risk, ethics, model validation, data protection. Some of that work will map onto the Act’s requirements. Much of it will not, because a role written without the regulation in view rarely produces the conformity assessments, technical documentation, and human-oversight structures the Act specifically demands. Readiness is even thinner than the headcount suggests. Simply counting governance hires overstates how many people are actually working the law. What job descriptions actually name The EU AI Act is visible in governance roles — but still absent from most job ads. Across the laws and frameworks most relevant to AI governance

AI Agents and Compliance

78% of organizations have no formal policies for creating or removing AI agent identities, according to a 2026 report from the Cloud Security Alliance and Oasis Security. The same research found that 92% are not confident that their legacy identity and access management tools can handle the risks agents introduce. Those two numbers describe the problem in full: enterprises are deploying autonomous software that reads email, queries databases, and triggers actions across production systems, and most of them cannot say who authorized it, what it can touch, or how they would prove any of that to an auditor. This is not a future problem. Agents are already operating inside regulated environments governed by the GDPR, HIPAA, SOX, and the EU AI Act. Every access decision an agent makes is a compliance event, whether or not anyone is logging it. This article covers what regulators actually expect, where traditional IAM falls short, and how to build an access framework for AI agents that survives an audit. Understanding the Compliance Landscape for AI Agents Key Regulations Impacting AI Agent Access No regulation says “AI agent” and then hands you a checklist. Instead, agents inherit obligations from every framework that governs the data and systems they touch. Under the GDPR, an agent processing personal data triggers the full set of principles in Article 5: lawfulness, purpose limitation, data minimization, and accountability. If an agent makes decisions that produce legal or similarly significant effects on individuals, Article 22 restrictions on automated decision-making apply as well. HIPAA requires covered entities to implement access controls, audit controls, and integrity protections for electronic protected health information under the Security Rule, and an agent with access to ePHI is subject to the same technical safeguards as a human workforce member. SOX demands that access to financial reporting systems be controlled, segregated, and reviewable, which becomes genuinely difficult when an autonomous agent can touch the general ledger. The EU AI Act adds an AI-specific layer, and its timeline is widely misunderstood. Following the Digital Omnibus agreement, obligations for standalone high-risk systems under Annex III were deferred to December 2, 2027. But the Article 50 transparency obligations still apply from August 2, 2026, meaning agents that interact with people in the EU must disclose their artificial nature on the original schedule. Treating the Omnibus as a blanket delay is one of the most common compliance mistakes being made right now. Important: The Digital Omnibus deferred the high-risk regime, not the whole Act. If an AI agent interacts with users in the EU, the August 2, 2026, transparency requirements were not moved, and the AI Office’s enforcement powers go live on the same date. Do not stand down 2026 workstreams based on headlines about the 2027 deferral. How AI Agents Create New Compliance Risks Agents break the assumptions most compliance programs are built on. A human user requests access, receives a role, and behaves within a predictable envelope. An agent reasons about its own goals, chains tool calls across systems, and can attempt actions its designers never anticipated. It operates at machine speed and machine volume, so a misconfigured permission produces thousands of non-compliant data touches before anyone notices. And because agents frequently run on shared service accounts or borrowed OAuth tokens, attribution collapses: the audit log says the CRM was queried, but not by whom, for what purpose, or under whose authority. The Gap Between Traditional IAM Compliance and Agentic AI Traditional IAM assumes identities are stable, access needs are predictable, and behavior maps to a job description. None of that holds for agents. A 2026 Cloud Security Alliance survey found that 68% of organizations cannot reliably distinguish AI agent activity from human activity in their logs. For a compliance function, that is disqualifying. If you cannot separate agent actions from human actions, you cannot certify access, demonstrate segregation of duties, or respond to a data subject access request with confidence. Core Compliance Requirements for AI Agent Access Auditability and Traceability of Agent Actions Every major framework converges on the same demand: show your work. For agents, a login timestamp is not enough. A defensible audit trail captures the full chain of custody for each action: which agent acted, which human or process delegated the authority, which tool or API was invoked, which data was accessed, and what the outcome was. Gartner’s 2026 Market Guide for what it calls “guardian agents” describes exactly this pattern of recording agent-to-tool-to-target chains for compliance reporting and incident response. Data Protection and Privacy Obligations Agents must operate inside the same data protection perimeter as everything else. That means Data Loss Prevention (DLP) controls apply to agent outputs, not just human uploads. It means an agent’s access to personal data needs a lawful basis, documented before deployment, not reverse-engineered after. And it means retention rules follow the data into whatever context window, vector store, or scratchpad the agent moves it into. Separation of Duties in Autonomous Systems Separation of duties exists so that no single actor can both commit and conceal an error or a fraud. A single agent granted permissions across procurement, approval, and payment reconstitutes exactly the toxic combination SOX controls were designed to prevent, except now it executes at machine speed. The control translates directly: no agent should hold permission sets that a human in the same process would be prohibited from combining, and multi-agent workflows need the same conflict analysis as human role assignments. Consent, Purpose Limitation, and Data Minimization Purpose limitation is the principle that agents most naturally violate. An agent given broad access “to be helpful” will use data collected for one purpose to accomplish another, because nothing in its architecture knows the difference. Compliance-ready agent access means scoping data access to the declared purpose of the task and enforcing that scope technically rather than hoping the system prompt holds. Insider Note: In practice, the purpose limitation failures we see are rarely dramatic. They look like a support agent enriching a ticket with data pulled from the sales