Enterprise data licensing, term by term.

The full sequence of steps an AI agent or person takes to complete a task, including actions, tool calls and results.

Information used to teach an AI model patterns, such as text, conversations, images, audio, code or action records.

A cap on how many requests a program can make to a software service in a given time.

When test examples leak into a model's training data, inflating its scores.

A written text version of a phone or video call, often produced by speech-to-text software.

A documented record of who held, accessed or changed something, and when, from collection to delivery.

A model's intermediate reasoning steps written out before its final answer.

Records of people operating software, such as screen recordings or click and keystroke sequences.

The tools and processes a company uses to collect, record and honor people's choices about how their data is used.

Adding labels or metadata to raw data so models can learn from it.

Irreversibly transforming data so individuals cannot be identified.

Treating a dataset as something designed, documented and maintained for its users, with clear ownership and quality standards.

A company that collects personal information about consumers, often from public and commercial sources, and sells it to others.

An inventory of a company's data assets with descriptions, owners and metadata so people can find and understand them.

A secure environment where parties can analyze combined data without seeing each other's raw records.

Under GDPR, the organization that decides why and how personal data is processed.

An organization where members pool data and govern its use collectively, often sharing the benefits.

Copying records out of a software system into files or another system.

Policies and controls that manage data's ownership, quality, access, use, retention and deletion.

An organization that connects data holders with data users and helps arrange access on agreed terms.

A list of the systems, record types, volumes and date ranges a company holds.

A data platform that combines low-cost data lake storage with database-style tables and governance.

Granting specific rights to a dataset while the supplier retains ownership of their original records and SourceX owns the deidentified version.

A map of where data came from and every step it passed through before reaching its current form.

Collecting, keeping and sharing only the data needed for a specific purpose.

A contract that sets the rules for how a processor handles personal data for a controller.

Under GDPR, an organization that handles personal data on behalf of a controller and only on its instructions.

The documented origin and history of a dataset: where it came from, who authorized it and what was done to it.

A requirement or commitment that data is stored and processed in a specific country or region.

The principle that data is subject to the laws of the country or region where it is collected or stored.

The person responsible for the quality, documentation and appropriate use of a dataset.

A legal structure in which independent trustees manage data on behalf of beneficiaries.

A defined collection of data assembled for a purpose, such as training or evaluating a model.

A short document describing a dataset: contents, sources, collection, preparation, intended uses and limitations.

Licensing a specific, defined dataset rather than ongoing access to a system.

A structured questionnaire, proposed by Gebru and colleagues in 2018, for documenting a dataset's motivation, composition, collection and uses.

Identifying who spoke when in an audio recording, separating speakers into labeled turns.

A mathematical technique that adds carefully measured noise to data or results so individuals cannot be singled out.

Video recorded from a person's point of view, usually with a head-mounted or body-worn camera.

Lists of numbers that represent the meaning of text, images or other data so similar items sit close together.

AI systems built for or deployed within organizations to perform business tasks.

Data an organization generates while operating: communications, transactions, documents, tickets, code and workflows.

Extract, transform, load: moving data out of source systems, cleaning and reshaping it, and loading it somewhere new.

A collection of examples used to measure model performance, kept separate from training data.

A license where the owner agrees not to license the same data to anyone else within an agreed scope.

The HIPAA method where a qualified statistical expert certifies that the risk of identifying individuals is very small.

A contract term limiting a license to specific purposes, products or industries.

Further training of an existing model on a smaller, focused dataset to improve it for a task or domain.

A carefully verified set of examples with trusted correct answers used as a benchmark.

Data deliberately kept out of training so it can test whether a model generalizes.

Data created by people rather than generated by models, including expert work products and human judgments.

A contract promise by one party to cover the other's losses if certain things go wrong, such as a third-party claim.

Training a model on instruction and response pairs so it follows requests well.

A privacy measure where each record is indistinguishable from at least k−1 others on identifying attributes.

A collection of articles, guides and answers a company maintains so customers or staff can solve problems themselves.

One of the six legal grounds under GDPR that permit processing personal data, such as consent, contract or legitimate interests.

A GDPR lawful basis that allows processing when a company's interest is balanced against, and not overridden by, individuals' rights.

A fixed amount a buyer agrees to pay regardless of how much revenue share or usage fees would otherwise produce.

Measuring how well an AI model performs on defined tasks.

A promise that a party will receive terms at least as good as those given to others.

Data combining more than one type, such as text with images, audio or video.

NER: software that finds and labels names of people, companies, places, dates and similar items in text.

A license that lets the owner license the same data to other buyers.

Optical character recognition: software that turns text in scans, photos or PDFs into machine-readable text.

Records of how a business runs day to day, such as work orders, tickets, approvals and process logs.

A choice that lets a person stop a company from using their data for a particular purpose.

Information that can identify a person directly or in combination, such as names, emails, account numbers, faces or voices.

Protected health information: health details tied to an identifiable person, such as diagnoses, treatment notes, claims or appointment records.

Automated scanning that finds personal details, such as names, emails, phone and account numbers, in data.

Training steps after pretraining that shape a model's behavior, such as instruction tuning and reinforcement learning.

Examples where people compare two or more model responses and pick the better one.

The first, large-scale stage of training a language model on broad text to learn general patterns.

Training that rewards each correct step in a model's reasoning, rather than only the final answer.

Data a company owns and controls that is not publicly available.

Replacing direct identifiers, such as names, with codes so records no longer point to a person without extra information.

Linking supposedly anonymous data back to a specific person, often by combining it with other information.

Data collected from actual activity rather than simulated or generated.

Deliberately testing an AI system by trying to make it fail, misbehave or reveal information it should not.

Removing or masking sensitive information, such as names, account numbers or addresses, from a record before it is shared.

A training approach in which a model learns by receiving rewards or penalties for its actions.

A company's rules for how long each type of record is kept and when it is deleted.

RAG: an AI technique where a model looks up relevant documents at answer time and uses them to write its response.

A payment model where the licensor receives a percentage of revenue tied to the licensed asset.

A model trained to score responses by how well they match human preferences or task success.

A simulated setting where an AI agent takes actions, receives feedback and learns through reinforcement learning.

Reinforcement learning from AI feedback: using an AI model's judgments in place of, or alongside, human feedback.

Reinforcement learning from human feedback: training a model using human preference judgments.

The HIPAA method that de-identifies health data by removing 18 specified identifiers and confirming no actual knowledge of re-identification.

Standard operating procedure: a written, step-by-step description of how a team does a recurring task.

Software that converts spoken audio into written text.

Template contract terms approved by the European Commission for transferring personal data from the EU to other countries.

Permission for a licensee to grant some of its rights to third parties.

Fine-tuning with labeled input and output examples showing the desired response.

Data generated by models or simulations rather than collected from real activity.

Records from help desk or issue tracking tools: requests, conversations, status changes, assignments and resolutions.

The number of tokens in a dataset or text, used to estimate size and cost.

The small chunks of text, often word pieces, that language models read and produce.

Examples of a model or person calling tools, such as search, APIs or calculators, and using the results.

A seller's promise that it owns, or has the right to license, what it is providing.