---
title: "Glossary"
description: "Plain-language definitions of the privacy and AI terms that matter when sensitive data meets generative models — masking, tokenization, PII, PHI, prompt leakage, and more."
url: "https://sonomos.ai/glossary/"
source: "https://sonomos.ai/glossary/"
---

# Glossary

Plain-language definitions of the privacy and AI terms that show up most often when people ask how Locke works. Each entry links to the related concepts and to the parts of the product that handle them.

- ## Data Masking

  Replacing sensitive values with realistic-looking substitutes so the data remains usable while the originals never leave the device.

  Read definition
- ## Tokenization

  Replacing a sensitive value with a non-sensitive token that maps back to the original through a separate, controlled lookup.

  Read definition
- ## Personally Identifiable Information (PII)

  Any information that can identify a specific person, either directly (name, SSN) or in combination with other data (zip code + birthdate).

  Read definition
- ## Protected Health Information (PHI)

  Individually identifiable health information held or transmitted by a covered entity, regulated under HIPAA in the United States.

  Read definition
- ## Prompt Leakage

  The unintended disclosure of sensitive information through the content of a prompt sent to an AI model.

  Read definition
- ## Zero-Knowledge Processing

  An architecture where the service provider has no technical ability to read user data, because all sensitive processing happens on the client.

  Read definition
- ## Privacy Layer for AI

  A control point between a user and an AI service that detects, transforms, or blocks sensitive data before it reaches the model.

  Read definition
- ## On-Device AI

  Running model inference or pre-processing entirely on the user's machine instead of sending data to a cloud service.

  Read definition
- ## Redaction

  Removing or blacking-out sensitive content so it is no longer present in the document or prompt.

  Read definition
- ## Contextual Matching

  Detecting sensitive data by combining structural patterns with surrounding context, instead of relying on regex alone.

  Read definition
- ## Prompt Injection

  An attack where adversarial instructions hidden inside user input or external content trick an AI model into ignoring its original instructions.

  Read definition
- ## Shadow AI

  Employees using AI tools at work without organisational sanction — the AI-era version of shadow IT.

  Read definition
- ## Data Loss Prevention (DLP)

  A category of security tooling that inspects outbound data flows to prevent sensitive content from leaving an organisation's control.

  Read definition
- ## Named Entity Recognition (NER)

  An NLP technique that identifies and classifies real-world entities — people, organisations, locations — within free-form text.

  Read definition
- ## Attorney-Client Privilege

  A legal doctrine protecting confidential communications between a lawyer and client from compelled disclosure — and one of the easiest privileges to inadvertently waive.

  Read definition
- ## HIPAA

  The U.S. Health Insurance Portability and Accountability Act, which sets privacy and security rules for individually identifiable health information held by covered entities.

  Read definition
- ## GDPR

  The EU General Data Protection Regulation — the world's most comprehensive data-protection law, governing how organisations handle personal data of EU residents.

  Read definition
- ## CCPA

  The California Consumer Privacy Act and its 2023 successor, the CPRA — the United States' most influential state-level privacy law.

  Read definition
- ## Business Associate Agreement (BAA)

  A HIPAA-mandated contract between a covered entity and a vendor that handles PHI on its behalf, defining each party's privacy and security obligations.

  Read definition
- ## Data Processing Agreement (DPA)

  A GDPR-mandated contract between a controller and a processor that governs how personal data is handled on the controller's behalf.

  Read definition
- ## Pseudonymization

  Replacing direct identifiers with stable pseudonyms so the data can no longer identify a person without additional information held separately.

  Read definition
- ## GLBA (Gramm-Leach-Bliley Act)

  The U.S. federal law requiring financial institutions to protect the nonpublic personal information of their customers and explain their data-sharing practices.

  Read definition
- ## FERPA (Family Educational Rights and Privacy Act)

  The U.S. federal law that protects the privacy of student education records held by institutions receiving federal funding.

  Read definition
- ## De-identification

  The process of removing or obscuring identifiers from health data so it no longer identifies an individual, removing it from HIPAA's scope.

  Read definition
- ## Data Minimization

  The privacy principle of collecting and processing only the minimum personal data necessary for a specific, stated purpose.

  Read definition
- ## PCI DSS

  The Payment Card Industry Data Security Standard — the global security framework that governs how organisations protect cardholder data wherever it is stored, processed, or transmitted.

  Read definition
- ## SOC 2

  A security audit framework for service organisations that tests whether controls over security, availability, processing integrity, confidentiality, and privacy are suitably designed and operating effectively.

  Read definition
- ## BIPA (Biometric Information Privacy Act)

  Illinois' 2008 law requiring informed written consent before collecting biometric data — the most-litigated AI-adjacent privacy statute in the United States, with statutory damages of $1,000–$5,000 per violation.

  Read definition
- ## Automated Decision-Making (ADM)

  Using algorithms or AI systems to make or substantially influence decisions about individuals — regulated under GDPR Article 22, Colorado SB 24-205, and a growing number of state and sector laws.

  Read definition
- ## EU AI Act

  The European Union's comprehensive AI regulation, which classifies AI systems by risk level and imposes conformity assessments, transparency obligations, and prohibitions on certain high-risk uses.

  Read definition
- ## COPPA (Children's Online Privacy Protection Act)

  The U.S. federal law requiring verifiable parental consent before online services collect personal information from children under 13.

  Read definition
- ## Anonymization

  Irreversibly transforming personal data so that no person can be identified, even with additional information.

  Read definition
- ## Data Residency

  A commitment about the geographic location where data is stored and processed — and, for AI tools, three separate questions: where data rests, where inference runs, and who can access it remotely.

  Read definition
- ## Standard Contractual Clauses (SCCs)

  European Commission-approved contract templates that provide a lawful mechanism for transferring personal data from the EEA to countries without an adequacy decision, such as the United States.

  Read definition

Looking for the deeper how-it-works view? See [how detection works](https://sonomos.ai/support/locke/how-it-works/) in the Support Center, the [security](https://sonomos.ai/security/) page for the architectural model, or the [comparisons](https://sonomos.ai/compare/) page for how Locke sits next to other approaches.
