Back to allhub.io
EU AI Act Art. 50 — Transparency Disclosure

AI Transparency

AllHub is an AI-powered platform. This page explains which AI systems we operate, how they work, what data they process, and your rights under GDPR and the EU AI Act.

Our role under the EU AI Act

AllHub operates as a Downstream Provider under Regulation (EU) 2024/1689 (EU AI Act). This means we build AI-powered applications on top of foundation models provided by third parties, and we deploy those applications to online store owners (Deployers) who use them to serve their shoppers (End Users).

AllHub

Downstream Provider

Builds and operates the AI platform and pipeline.

Store Owners

Deployers

Configure and deploy AllHub agents in their stores.

Shoppers

End Users

Interact with the AI agent inside the store.

AI systems we operate

We operate four AI systems. None are classified as High Risk under the EU AI Act.

agent-sales-v1

Cortex Sales Agent

Generative AI — Conversational

Transparency Risk

Conversational AI assistant deployed in online stores. Answers buyer questions, recommends products and guides shoppers to checkout.

Transparency disclosure (Art. 50)

You are being assisted by an AI agent.

Known limitations

  • May produce inaccurate product information if catalog is out of date
  • Does not have access to payment data
  • Intent matching confidence threshold: 0.50

Not designed for

  • Credit or financing decisions
  • Medical advice
  • Legally binding content
Human escalation path available

store-brain-v1

Store Brain

Generative AI — Knowledge Retrieval

Minimal Risk

Internal knowledge system that synthesizes store data — conversations, catalog gaps and funnel signals — to answer store-owner questions.

Transparency disclosure (Art. 50)

Internal tool for store owners only. Not visible to shoppers.

Known limitations

  • Answers are based on aggregated store data; not real-time
  • Accuracy depends on the quality of ingested data

Not designed for

  • Customer-facing decisions
  • Financial projections

demand-simulator-v1

Demand Simulator

Predictive AI

Minimal Risk

Simulates how buyers will react to store changes (price updates, product launches, campaigns) using historical store data and AI reasoning.

Transparency disclosure (Art. 50)

Simulation output is advisory only. Final decisions are made by the store owner.

Known limitations

  • Projections are probabilistic, not guaranteed
  • Accuracy depends on store data volume and recency

Not designed for

  • Automated pricing without human review
  • Individual customer profiling
Human escalation path available

decision-extractor-v1

Decision Extractor

Generative AI — Analysis

Transparency Risk

Analyzes buyer conversation turns to extract purchase intent signals and decision patterns. Used to improve the store's AI responses.

Transparency disclosure (Art. 50)

Conversation analysis is performed on pseudonymised session data only.

Known limitations

  • Analysis is aggregate — no individual profiling
  • SHA-256 pseudonymisation applied to all session identifiers

Not designed for

  • Individual user tracking
  • Building personal profiles

Data protection measures

GDPR Art. 35 — Privacy by Design

L0 Shield DLP

All buyer input is scanned for personal data (email, phone, ID numbers, payment data) before reaching the AI. Detected PII is blocked — never processed by the model.

SHA-256 Session Pseudonymisation

Session identifiers are hashed with SHA-256 before storage. The raw session ID is never persisted in any database.

6-Month Data TTL

Conversation audit logs and decision traces are automatically deleted after 6 months. No long-term retention of buyer interaction data.

Tenant Isolation

Each store's data is fully isolated under its own tenant namespace. No cross-tenant data access is possible by design.

Decision Trace Audit Log

Every AI response is logged with the reasoning path (intent matched, layers invoked, confidence score) for full EU AI Act Art. 14 traceability.

Kill Switch per Agent

Each AI agent can be disabled instantly by the store owner or platform administrators without any code deployment.

FAST Compliance Score

Fairness · Accountability · Security · Transparency

We assess our AI pipeline monthly against four pillars. Current score: 9.75 / 10 — COMPLIANT

10/10

Fairness

10/10

Accountability

9/10

Security

10/10

Transparency

Your rights

GDPR Chapter III — Rights of the data subject

Right to Access: Request a copy of the data we hold about your interactions.
Right to Rectification: Request correction of inaccurate personal data.
Right to Erasure: Request deletion of your data (right to be forgotten).
Right to Portability: Receive your data in a machine-readable format.
Right to Object: Object to the processing of your data for direct marketing.

To exercise any of these rights, contact us at privacy@allhub.io. We respond within 30 days.

Regulatory basis: EU AI Act (Regulation 2024/1689) · GDPR (Regulation 2016/679) · Deeploy AI Governance Framework v2.0.
AllHub operates in the EU (europe-west1). Data is not transferred outside the European Economic Area.
Last updated: April 2026. Next review: July 2026.