AI GOVERNANCE COMMAND CENTRE · GUARDLENS

Trustworthy AI isn't an accident. It's a practice. Bias. Risk. Compliance. Explained.

GuardLens is the governance layer your AI stack is missing — register models, detect bias, explain decisions, score compliance, monitor models live in production, and stay ahead of EU AI Act enforcement. All in one platform.

GDPR COMPLIANT NO CREDIT CARD FREE TRIAL · 1 SYSTEM IRELAND · EU-BASED
6
GOVERNANCE
MODULES
4
CLOUD
CONNECTORS
€35M
MAX AI ACT
FINE
Aug '26
ENFORCEMENT
DEADLINE
€0
TO GET
STARTED
--
DAYS
--
HOURS
--
MINS
--
SECS
⏳ until EU AI Act enforcement — August 2026
AI RegistryRisk ScoringBias & FairnessExplainabilityCloud ConnectorsAudit LogEU AI Act · Art. 5Annex III · High RiskArt. 9 · Risk MgmtArt. 12 · Record KeepingArt. 13 · TransparencyArt. 14 · Human OversightHuggingFace · LiveAWS SageMaker · LiveAzure ML · SoonVertex AI · SoonAI Agent Scanning · NewCode Scanner · New AI RegistryRisk ScoringBias & FairnessExplainabilityCloud ConnectorsAudit LogEU AI Act · Art. 5Annex III · High RiskArt. 9 · Risk MgmtArt. 12 · Record KeepingArt. 13 · TransparencyArt. 14 · Human OversightHuggingFace · LiveAWS SageMaker · LiveAzure ML · SoonVertex AI · SoonAI Agent Scanning · NewCode Scanner · New

THE COST OF DOING NOTHING

Your AI is making decisions.
Who's checking the work?

From August 2026, the EU AI Act is enforceable. Regulators can fine companies up to €35M for non-compliance — and most AI teams have no idea where they stand. GuardLens changes that in minutes, not months.

€35M
or 7% global turnover
for prohibited AI practices
€15M
or 3% global turnover
for high-risk AI violations
€7.5M
for incorrect information
given to regulators
€0
to start with GuardLens
and know where you stand
THE REAL DEAL
Most AI governance tools are spreadsheets in disguise.

Rename a Google Sheet, add a logo, charge €50K — that's the playbook. GuardLens is a live, automated platform. Everything that matters happens without you lifting a finger.

TYPICAL "GOVERNANCE PLATFORM"
You manually fill in risk fields — in a form that's really just a database table
Compliance score is a number you type in yourself
Bias detection means a checklist — no actual statistical testing
"Monitoring" is a column you update manually every quarter
Audit log is whatever someone remembered to write down
Gets stale within a week. Useless in a regulator audit.
GUARDLENS — ACTUALLY AUTOMATED
Models auto-classified against Annex III & Art. 52 the moment they're registered
Risk scores computed automatically — weighted across compliance, fairness & documentation
Bias detection runs real statistical tests — DIR, DPD, EOD across demographic groups
Live monitoring fires alerts the moment drift or fairness thresholds are breached in production
Immutable audit trail — every action logged automatically, timestamped, regulator-ready
Always current. Export a full compliance report in one click, any time.
MISCONCEPTION #1

A form you fill in is not a governance platform. If your "compliance score" is a number you typed in yourself, you don't have governance — you have a spreadsheet with a login screen.

MISCONCEPTION #2

Knowing which AI tools your team uses is a starting point — not a governance strategy. Real governance means understanding what your models are actually doing, who they're affecting, and being able to prove compliance when it matters.

This is what a real AI governance platform looks like. Built to satisfy regulators, not slide decks.

HOW IT WORKS
Up and running in three steps.

No lengthy onboarding. No professional services engagement. Register your first AI system and see your governance posture in under 5 minutes.

Register your AI systems
Connect your cloud ML platforms or add models manually. GuardLens auto-classifies each one against Annex III and Article 52 of the EU AI Act the moment it's registered.
01 · REGISTER
Score, explain & detect bias
Run a Smart Score assessment to get your weighted compliance score. Check for proxy features and bias across demographic groups. See exactly what's driving your model's decisions.
02 · ASSESS
Govern, audit & export
Every action is logged in an immutable audit trail. Export a full PDF compliance report in one click — ready for regulators, investors, or internal governance reviews.
03 · GOVERN
EU AI ACT RISK TIERS
Where does your AI actually sit?

The EU AI Act classifies every AI system into one of four risk tiers. Your obligations — and the fines for non-compliance — depend entirely on which tier you're in. Click each to explore.

UNACCEPTABLE
HIGH RISK
LIMITED
MINIMAL
BANNED
Unacceptable Risk
These AI systems are prohibited outright under Article 5 of the EU AI Act
🚫
Social scoring
Government or private systems that score citizens based on social behaviour or personal characteristics
🎭
Subliminal manipulation
AI that exploits psychological vulnerabilities to manipulate behaviour without conscious awareness
👁️
Real-time biometric surveillance
Live facial recognition in public spaces by law enforcement (with narrow exceptions)
WHAT THIS MEANS FOR YOU
If your AI falls here, it must be discontinued. There is no compliance path — only removal.
Penalty for violation: up to €35M or 7% of global turnover
HIGH RISK
High Risk — Annex III
Significant compliance obligations under Articles 9–15. Most enterprise AI falls here.
👔
HR & Recruitment AI
CV screening, candidate scoring, interview analysis tools used in hiring decisions
💳
Credit Scoring
AI systems that determine creditworthiness, loan eligibility, or insurance pricing
🏥
Medical Diagnosis
AI assisting clinical decisions, diagnosis, or treatment recommendations
YOUR OBLIGATIONS
Risk management system documented and maintained throughout the lifecycle
Data governance — training data documented, biases assessed (Art. 10)
Technical documentation and audit logs kept (Art. 11, 12)
Human oversight mechanisms in place (Art. 14)
Accuracy, robustness and cybersecurity measures (Art. 15)
LIMITED RISK
Limited Risk — Article 52
Transparency obligations only. Users must know they're interacting with AI.
💬
Chatbots
Customer service bots, virtual assistants, and conversational AI that interact with people
🎨
Deepfake generators
AI that generates synthetic images, video or audio of real people — must be labelled
🤖
Emotion recognition
Systems that infer emotional states — users must be informed they are being analysed
YOUR OBLIGATIONS
Disclose to users that they are interacting with an AI system (Art. 52)
Label AI-generated content clearly
No registration or conformity assessment required
MINIMAL RISK
Minimal Risk
No specific obligations under the EU AI Act. Good governance still recommended.
🎮
AI in games
Spam filters, AI-powered game NPCs, and entertainment recommendation systems
📦
Product recommenders
E-commerce recommendation engines and personalisation tools with no significant impact on rights
🔍
Search & SEO tools
Content ranking, search optimisation, and non-sensitive classification systems
BEST PRACTICE (VOLUNTARY)
Register systems in GuardLens for full governance visibility even if not legally required
Document model decisions — regulators may still request this during investigations
Risk tier can change if use case evolves — monitor regularly
WHY GUARDLENS
How does it stack up against the alternatives?

There are three ways to approach AI governance. Here's an honest comparison.

Manual / Spreadsheets Hire a Consultant ⚡ GuardLens
Time to first assessment Weeks Weeks–months Under 5 minutes
Auto risk classification ✗ Manual ~ Sometimes ✓ Automatic
Bias & fairness detection ✗ Not included ~ Extra cost ✓ Built in
Explainability ✗ Not included ~ Extra cost ✓ Built in
Immutable audit trail ✗ Version control only ✗ Not provided ✓ Automatic
PDF compliance reports ~ Manual effort ✓ Provided ✓ One click
Cloud connector sync ✗ Manual entry ✗ Not provided ✓ Live sync
Stays up to date with regulations ✗ Your responsibility ~ Periodic reviews ✓ Continuous
Starting cost Staff time only €5,000–€50,000+ €0 — free tier
THE PLATFORM
Six modules. One command centre.

Everything AI governance needs — from first registration to regulator-ready audit reports — fully integrated and live from day one.

AI Registry & Risk Scores

Register every AI system and auto-classify against Annex III and Article 52 on entry. Weighted compliance scoring across fairness and documentation dimensions with full score history.

Annex III · Art. 9 · Art. 52

Bias & Fairness

Test models across demographic groups. Detect protected attribute leakage — gender, race, age. Surface disparate impact before it becomes a liability.

Art. 10 · Art. 15

Explainability

Feature importance analysis, proxy feature detection, what-if analysis, individual prediction breakdowns, and AI-written compliance narratives.

Art. 13 · Art. 14

Continuous Monitoring

Track model performance in production. Ingest snapshots of accuracy, drift, and fairness metrics. Auto-generate alerts when thresholds are breached — open, acknowledge, resolve.

Art. 72 · Post-Market

Connectors & Scanners

Live sync from major cloud ML platforms. AI agent scanning included — detect and register autonomous agents across your codebase and cloud infrastructure.

Models · Agents

Audit Log

Immutable timestamped record of every action. Filter by Create, Update, Delete, Score. Export PDF compliance reports in one click.

Art. 12 · Art. 17
EXPLAINABILITY ENGINE
Why is your model making that call?

Most AI teams can't answer that question. GuardLens surfaces exactly what drives your model's outputs — and flags hidden proxy features that could expose you to discrimination claims before a regulator finds them first.

Proxy feature detection — flags inputs like postcode or salary history that silently encode protected attributes
What-if analysis — change any input and instantly see how the prediction shifts
AI compliance narratives — plain-English model explanations ready to attach to audit reports
Notebook export — download runnable explainability analysis code for your data science team
EXPLAINABILITY · HR CV Screener
PROXY FEATURES DETECTED
2 input features may encode protected characteristics — review recommended
TOP FEATURE DRIVERS
Input A
High
Input B
High
Input C proxy
Med
Input D
Low
Input E proxy
Low
CONTINUOUS MONITORING
Know when your model starts to drift.

EU AI Act Article 72 requires post-market monitoring for all high-risk AI systems. GuardLens automates it — ingesting production snapshots, tracking fairness in deployment, and firing alerts the moment a metric breaches your configured threshold.

Performance snapshots — submit accuracy, F1, precision, recall, prediction drift, and feature drift per monitoring period
Fairness in production — track DIR, DPD, and EOD per sensitive attribute after deployment, not just at training time
Automatic alerts — configurable thresholds per org (DIR floor 0.80, drift ceiling 0.05) with high/medium severity and open → acknowledged → resolved workflow
Trend analysis — plot accuracy, F1, and fairness metrics over time across the last N snapshots
MONITORING · Loan Approval Model
ALERT · DIR BELOW THRESHOLD
Disparate Impact Ratio 0.76 < 0.80 minimum — fairness review required
LATEST SNAPSHOT METRICS
Accuracy
82%
F1 Score
79%
Prediction Drift
0.06
DIR (gender)
0.76
DPD (age)
0.08
CONNECTORS & SCANNERS
Your models and agents, wherever they live.

Connect once. GuardLens syncs models and scans for AI agents across your cloud and codebase — auto-classifying everything for EU AI Act compliance.

ML PLATFORMS
ML Platform A
● CONNECTED
Auto-import models from any namespace. Live sync on registration.
ML Platform B
● CONNECTED
Secure cloud architecture. Zero config after setup.
Azure ML
● CONNECTED
Import models from Azure Machine Learning workspaces.
Vertex AI
● CONNECTED
Import models from Google Cloud Vertex AI registry.
AI AGENTS & SCANNERS NEW
Agent Connectors
○ Configure
Scan and import AI agents from major cloud platforms — auto-classified for compliance.
Code Scanner
● CONNECTED
Scan your codebase for AI agent patterns — detects agent frameworks across repos.
DEVELOPER API
Governance baked into your pipeline.

The GuardLens API lets you automate compliance checks, register models programmatically, and gate deployments on governance scores — directly from your CI/CD pipeline.

CI/CD integrationGate deployments on governance scores automatically
Auto-registrationRegister models programmatically the moment they're trained
Real-time scoresPull live compliance scores into your own dashboards
Audit on demandPull audit logs into your SIEM or data warehouse
API Access — Early Release

Full REST API documentation is available to approved partners and enterprise teams. Request access to receive your API key and onboarding guide.

Available on Professional and Enterprise plans
WHAT TEAMS SAY
Built with teams who ship AI for real.
★★★★★
"Finally a compliance tool that doesn't need a law degree. We went from zero governance to full AI Act coverage in a week."
Head of AI, FinTech startup
Series A · 40 employees
★★★★★
"The bias detection caught a proxy feature in our HR model we'd completely missed. That alone justified the whole platform."
Lead Data Scientist
Healthcare AI company
★★★★★
"We passed our EU AI Act audit on the first attempt. The audit trail and compliance narratives made the whole process straightforward."
CTO, Enterprise SaaS
Series B · 200 employees
GET ACCESS
Start free.
Scale when ready.

Start with a free trial — no credit card needed. Register your first AI system, run a full compliance assessment, and see your governance posture in under 5 minutes.

NO CREDIT CARD · FREE TRIAL · 1 SYSTEM · LIMITED TIME
Free Trial — Explore & Demo
1 AI system, full compliance assessment, audit log included. Free for a limited time.
€0 · limited offer
Early Access — Teams
Full platform access during beta. All 6 modules, cloud connectors, PDF reports, priority onboarding.
Discounted early-access pricing — contact us
Enterprise & Consulting
Custom deployment, AI Act gap analysis, responsible AI workshops, and bespoke model audits.
Custom — Contact
WORKSHOPS
Learn responsible AI. By doing it.

Hands-on workshops for every team — from executives to engineers. We help organisations move beyond buzzwords and build AI that is fair, transparent, and aligned with human values.

01
Introduction to Responsible & Ethical AI
Leaders & cross-functional teams
A foundational workshop covering the core principles of responsible AI — fairness, transparency, accountability, and privacy. Participants explore real-world examples of ethical failures and understand why responsible practices are critical for building trust and long-term value.
Book this workshop →
02
AI Risk Assessment & Impact Analysis
Product, compliance & risk teams
Helps organisations proactively identify, assess, and mitigate risks associated with AI systems. Teams gain practical frameworks for conducting ethical impact assessments and integrating risk-awareness into development and deployment workflows.
Book this workshop →
03
Bias & Fairness in Machine Learning
Data scientists & ML engineers
A hands-on technical workshop exploring how bias enters datasets and algorithms, and how to detect, measure, and reduce it. Participants use practical tools and real-world examples to build fairer, more inclusive AI models.
Book this workshop →
04
Explainability & Transparency in AI
Technical teams & product managers
Focuses on making AI systems more transparent and interpretable. Participants learn how to apply explainability tools to communicate model behaviour clearly and responsibly, enabling better user trust and regulatory compliance.
Book this workshop →
05
AI Policy, Regulation & Compliance Readiness
Legal, compliance & executive teams
An overview of the evolving global regulatory landscape, including the EU AI Act. Helps organisations assess where they stand, understand how regulations apply to their AI use cases, and build internal policies aligned with ethical and legal standards.
Book this workshop →
06
Ethical AI by Design
Product managers, UX & developers
A practical workshop integrating ethical thinking into the product design and development process. Through real scenarios and inclusive design principles, teams learn to build AI systems that prioritise user well-being, safety, and fairness from day one.
Book this workshop →

All workshops are available in-person or remote. Custom formats available for enterprise teams.

FOLLOW ALONG
Stay in the loop on AI governance.
LinkedIn X / Twitter YouTube
AI EDGE
Fresh reads & podcasts.
23 MAY 2025
Is AI Booming Exponentially?
Read →
15 SEP 2025
Navigating Trustworthy AI for Secure Systems
Read →
4 JUN 2025
What is Responsible and Ethical AI?
Read →
Read all articles on Hashnode →

Trustworthy AI isn't an accident.
It's a practice.

Start governing your AI systems today — free for your first model, no credit card required.

NO CREDIT CARD · FREE TRIAL FOR 1 SYSTEM · GDPR COMPLIANT
FAQ
Common questions, straight answers.
The EU AI Act is the world's first comprehensive AI regulation. It classifies AI systems into four risk tiers and sets compliance obligations for each. Enforcement begins August 2026, with fines of up to €35M or 7% of global turnover for the most serious violations.
Yes. The EU AI Act applies to any organisation whose AI systems are used in the EU — regardless of where the company is based. If your product or service reaches EU users, you are in scope.
It depends on your use case. HR tools, credit scoring, medical AI, and recruitment software typically fall into the High Risk tier (Annex III). Customer chatbots and content generators fall under Limited Risk (Article 52). GuardLens auto-classifies your system the moment you register it.
Consultants are expensive (€5,000–€50,000+), slow, and deliver a point-in-time report. GuardLens gives you continuous, automated compliance monitoring, real-time risk scoring, bias detection, and audit-ready reports — starting free, with no wait time.
Yes — GuardLens offers a free trial for your first AI system, with no credit card required. You get full access to compliance assessment, audit logging, and risk scoring. The free trial is available for a limited time. Teams and enterprise plans are available for larger organisations.
Click "Request Access" and fill in a short form — we'll be in touch within one business day to get you set up. The free trial covers your first AI system with full compliance assessment, risk scoring, and audit logging included.