August 2, 2026 — EU AI Act full compliance deadline · 4 months remaining
EU AI Act Compliance

Do you know which of your AI systems are high-risk?

Klarr maps your AI inventory against EU AI Act obligations. We identify what is prohibited, what is high-risk, and what needs to be documented — before the deadline.

See how it works
4
Months to deadline
€35M
Maximum fine per violation
+50%
Of enterprises unprepared
8
Compliance dimensions assessed
🇩🇪
Hetzner Germany
Infrastructure hosted on German-regulated servers. No US data routing.
🇫🇷
Mistral AI France
AI inference via European-native models. No dependence on US hyperscalers.
🇩🇰
Cordero Management Denmark
Operated from Copenhagen. EU jurisdiction. Legal and advisory accountability here.
🇪🇺
GDPR Native
Designed from the ground up under EU data protection law. No bolt-on compliance.
01

How it works

Step 01

Inventory your AI systems

Walk through a structured intake for each AI system in use across your organisation. Name, vendor, function, decision scope, and internal ownership — captured once, referenced always.

Step 02

Classify risk and obligation

Klarr maps each system against the EU AI Act annexes. Prohibited, high-risk, GPAI, or minimal — with the specific articles and documentation obligations that apply to your situation.

Step 03

Track and report readiness

Monitor progress across eight compliance dimensions. Export board-ready reports. Assign tasks to owners. Know your readiness stage at any moment — and what it takes to advance.

Live product preview
What your compliance dashboard looks like
Matas Group
Dashboard AI Systems Tasks Reports Settings
Current stage
Unaware
Next: Inventoried — begin AI system intake
AI Systems
0
identified
High-risk
—
classified
Tasks open
—
assigned
Days to deadline
116
Aug 2, 2026
System name Vendor Risk tier Stage Open tasks

This is what a CISO sees on day one. Your data, your systems, your obligations — in one place.

02

Eight dimensions of compliance

01 Risk Classification Map every system to its correct EU AI Act category. + See what this involves
Structured mapping against Annex III categories — Prohibited, High-risk, GPAI, or Minimal — with justification documented per system. Determines every other obligation that follows.Art. 6 · Annex III
02 Data Governance Training data documentation, bias testing, quality controls. + See what this involves
For high-risk systems: training data provenance, bias and discrimination testing, data quality assessment, and documentation of data preparation practices.Art. 10
03 Technical Documentation System design records, performance benchmarks, architecture. + See what this involves
Mandatory technical file covering general description, design specifications, development methods, validation and testing results, and monitoring procedures.Art. 11 · Annex IV
04 Transparency Disclosure obligations, user information, AI labelling. + See what this involves
Requirements to inform users when interacting with AI systems, disclose capabilities and limitations, and label AI-generated content. Applies broadly — not just to high-risk systems.Art. 13 · Art. 50
05 Human Oversight Override mechanisms, human-in-the-loop controls, escalation. + See what this involves
High-risk systems must be designed to allow human intervention, override, or shutdown. Deployers must assign individuals responsible for oversight and establish escalation pathways.Art. 14
06 Accuracy & Robustness Performance monitoring, error tracking, adversarial resilience. + See what this involves
Ongoing performance benchmarking, error rate tracking, cybersecurity robustness assessment, and resilience testing against adversarial inputs for high-risk systems.Art. 15
07 Conformity Assessment Self-assessment or third-party audit readiness for CE marking. + See what this involves
High-risk systems require a conformity assessment before market placement — either self-assessed or via a notified body, depending on the category. CE marking follows successful assessment.Art. 43
08 Post-Market Monitoring Incident reporting, serious risk escalation, surveillance. + See what this involves
Continuous monitoring after deployment. Serious incidents and near-misses must be reported to national authorities. Deployers must maintain a post-market monitoring plan and document findings.Art. 72 · Art. 73
03

Readiness stages

Stage 01UnawareNo inventory. No classification. Exposure unknown and unmanaged.
Stage 02InventoriedAI systems identified and documented. Risk tiers not yet assigned.
Stage 03AssessedRisk classification complete. Obligations identified per system.
Stage 04ControlledDocumentation, oversight, and monitoring mechanisms in place.
Stage 05CompliantAll obligations met. Audit-ready. Board-reportable. Deadline secure.
04

Pricing

What you are replacing

EU AI Act compliance is not a one-time legal review. It requires continuous inventory management, classification, task ownership across business units, board reporting, and post-market monitoring — for every AI system you deploy.

Most enterprises approaching this today are doing so with a combination of external consultants, legal hours, and internal coordination that was never designed for this regulation. That approach costs more, takes longer, and produces documentation that requires yet another round of review to be audit-ready.

Klarr replaces that process. The cost comparison is not with other SaaS tools. It is with the alternative.

Cost of the alternative
External consultant — initial AI assessment3–5 days at €1,500–2,500/day. No ongoing updates included.
€7,500–12,500
Legal review — one high-risk systemPer system, per annual update cycle.
€3,000–6,000
CISO coordination overhead8–15 hrs/month across inventory, tasks, and board prep.
€1,200–2,400/mo
Klarr ProfessionalUnlimited systems. All 8 dimensions. Board export. Advisory included.
€799/mo

Based on EU market consulting day rates. Your costs will vary. The direction will not.

Starter
€299
per month

For organisations beginning their AI Act compliance journey with a focused system inventory.

  • Up to 5 AI systems
  • Risk classification
  • Task management
  • PDF export
  • Email support
Professional
€799
per month

For enterprise compliance teams managing multiple systems across business units.

  • Unlimited AI systems
  • All 8 compliance dimensions
  • Board report export
  • Multi-user access
  • Article-level guidance
  • Dedicated advisory
Enterprise
Custom
bespoke engagement

For complex organisations requiring tailored assessment, workshop delivery, and ongoing advisory.

  • On-site assessment
  • Legal review integration
  • Executive briefings
  • Consortium support
  • SLA-backed advisory

August 2 is closer than your legal team thinks.

Request access or speak directly with our compliance team.