Glossary

AI and data-protection glossary

Check the meaning of AI and data-protection terms before you compare services or plan your project.

ADEGHOPRS
Audit trail
A tamper-evident log of system actions, prompts, responses and decisions, used for accountability and compliance.
Data residency
The location where data is stored or processed. Check storage, inference, logs, backups and support access separately.
DPA (data processing agreement)
The contract GDPR Article 28 requires between a data controller and a processor that handles personal data on its behalf.
EU AI Act
Regulation (EU) 2024/1689, the EU law for AI. Duties depend on your role and the system’s use, including prohibitions, high-risk controls, transparency and rules for general-purpose AI models.
GDPR
General Data Protection Regulation (EU) 2016/679. It governs how the personal data of people in the EU is collected, processed, stored and transferred.
Human-in-the-loop
A workflow where a person reviews, approves or corrects AI output before it takes effect.
On-prem deployment
Running software on infrastructure the customer controls, rather than on a vendor's cloud.
PII redaction
Automatic detection and masking of personally identifiable information (names, IDs, phone numbers, health data) before text is stored or sent to a model.
RAG (retrieval-augmented generation)
A pattern where an LLM answers from documents pulled out of your own knowledge base at query time, so answers can cite their sources and stay current.
Speaker diarisation
Grouping speech segments by speaker. Speaker labels do not by themselves identify a person or guarantee correct attribution.

Build your compliant AI or software project

Tell us your workflow and compliance requirements. We reply within four business hours and usually prepare a scope and architecture document in about five days.