Clear definitions for the terms used across the CIP standard, certification paths, and related materials. Start here if you want a simpler understanding of the language used across provenance, rights, consent, governance, and recognition.
An AI system that operates autonomously to take actions, make decisions, or produce outputs without continuous human oversight. In the CIP framework, agentic systems create distinct ingestion and output scenarios where subsisting rights may be engaged without human review.
The direction of rights movement from creative content into AI systems — covering training data, fine-tuning data, prompt inputs, and style or likeness. See also: Derivative flow.
An assessment conducted by an independent auditor reviewing live systems, documentation, and governance processes, rather than a knowledge examination. Used in Agency and Platform certification tracks.
Rights protecting faceprints, voiceprints, and other biometric identifiers. Distinct from NILP rights, biometric rights address the use of measurable biological data in AI model training and inference.
Creative AI Risk Index. The risk scoring framework used by CIP-certified underwriters to quantify exposure from subsisting rights failures at each stage of the AI pipeline.
Core Data Record. A structured record that captures key information about a creative asset, including the rights, consent, and provenance information attached to it. The CDR is the primary provenance infrastructure for AI pipeline entry.
A recognised mark awarded to an organisation — agency or platform — that has met the relevant organisational certification standard within the CIP framework. Distinct from individual credentials such as badges and designations.
A role-specific route within CIP with its own requirements, assessments, and outputs. The five paths are Creator, Agency, Legal Practitioner, Underwriter, and Platform.
The documented trail showing the movement of a creative asset through the rights chain — from original creator through transformations, ingestions, distributions, and uses. The CDR system and Rights Registry provide the infrastructure for chain of custody in the CIP framework.
A formal professional opinion issued by a CIP Legal Practitioner confirming that a contract, platform, or process meets the requirements of the CIP framework. Only practitioners holding the CIP Legal Practitioner Designation may issue this opinion.
The set of permissions, limits, and controls that define how a creative work may be used. The Consent Framework captures whether consent has been granted for AI training, fine-tuning, commercial use, and related activities, and on what terms.
Ongoing professional learning required to maintain certain CIP certifications. The Legal Practitioner Designation requires 10 hours of CPD per year. CPD activities must relate to developments in AI law, IP regulation, or CIP framework updates.
The direction of rights movement from AI systems into outputs — covering generated content, derivative works, synthetic media, and multimodal assets. See also: Attribution flow.
A formal professional title awarded within a CIP certification path. The CIP Legal Practitioner Designation and CIP Underwriting Designation are the two designation-level credentials in the framework.
A structured record of the provenance chain for a creative asset, showing each transformation, ingestion, and use event. The Evidence Graph supports audit and enforcement by providing a traceable history of rights movement.
Generative IP Liability. The insurance architecture within the CIP framework, covering liability arising from subsisting rights failures in generative AI content pipelines. GIPL policies are issued by CIP-certified underwriters.
A classification of the type of licence under which a creative work is held or shared. Classes include: Owned, Licensed (Commercial), Licensed (Non-Commercial), Open (Permissive), Open (Conditional), Restricted, Prohibited, and Training Ingestion.
Single-mode creative content within a single medium — an audio recording, a photograph, a written text, a film, a sculpture, a software application as a self-contained whole. Each medium has its own established rights regime, mature legal infrastructure, and well-understood category structure. A musician's song is media. A photographer's image is media. An author's chapter is media. Rights subsist in each as a discrete work, and the rights bundle is generally well-mapped to the medium type. In the CIP framework, content classified as `Single-Media` (per `CIP-Media-Class`, Enum 5 of CIP Classifications v1.0) is the default category. The framework's standard tools — `cip.md` declarations, CDRs, and the seven existing role-specific visual models — were designed primarily with single-media content in mind. The framework defaults to this classification when no other classification is asserted; the burden is on rights holders who hold multimedia or multimodal-media content to assert the higher-complexity classification explicitly.
The sensory or computational mode through which information is expressed or received. Common modalities are visual (still image), audiovisual (moving image with sound), audio (sound only), text (written language), structured data (machine-readable records), code (executable instructions), and three-dimensional (sculptural, architectural, virtual environment). A modality is a property of expression form, not of content meaning — the same idea can be expressed in multiple modalities. The term is distinct from "medium" (which refers to the physical or technological substrate of expression — a vinyl record is a medium, audio is the modality it carries) and from "format" (which refers to the encoding of a particular instance — MP3 is a format, audio is the modality). A multimodal-media system is one that operates across more than one modality at input, output, or both.
Rights protecting the authorship integrity of a creative work. They include the right of attribution (to be identified as the author), the right of integrity (to object to distortion or mutilation), and sometimes the right of disclosure or withdrawal.
Content that combines multiple media types into a single bundled work where the bundling itself produces a new aggregated right that subsists in the combination. A music video is multimedia (audio + film + sometimes embedded text + choreography). A video game is multimedia (visual art + sound design + music + code + narrative + sometimes performance capture). An interactive ebook is multimedia (text + image + audio + sometimes video). A theatrical production captured on film is multimedia (script + performance + music + design + cinematography). The legal complexity of multimedia is the *layered rights stack*. The underlying components each have their own rights, and the bundle has its own rights on top, often held by a different party from any of the component rights holders. A music video's master recording is owned by a label; the underlying composition by a publisher; the visual footage by a production company; the choreography by a choreographer; and the assembled audiovisual work itself by yet another party. AI ingestion of a music video engages all of those rights simultaneously. In the CIP framework, content classified as `Multimedia` (per `CIP-Media-Class`, Enum 5 of CIP Classifications v1.0) requires either an enumerated list of component rights or an attached layered-rights schedule. A `cip.md` declaration with `CIP-Media-Class: Multimedia` and no component-rights enumeration is treated as incomplete — the declarant has asserted the classification but has not made the layered rights legible. Vendor contracts using the CIP Vendor Representation should specify which media class the vendor's content falls into, and where that class is `Multimedia`, should reference the layered-rights schedule by name.
Content that exists not as a fixed bundled work but as a *generative or interactive system* that produces media outputs across multiple modalities in response to inputs. A trained AI model is multimodal media. A text-to-image generator is multimodal media. A voice-cloning system that produces audio outputs from text prompts is multimodal media. An interactive virtual environment is multimodal media. An agentic system that chooses outputs across text, image, audio, and code is multimodal media. The legal question here is fundamentally different from the multimedia case. For multimedia, the question is "which rights subsist in this fixed bundled work?" For multimodal media, the questions are "which rights subsist in the *outputs* of a generative system?", "which rights subsisted in the inputs that trained the system?", and "what relationship exists between the training inputs and the generative outputs?". CDPA 1988, EU AI Act Articles 50 and 53, and DUA Act 2025 are all currently working through these questions; the law is not settled. In the CIP framework, content classified as `Multimodal-Media` (per `CIP-Media-Class`, Enum 5 of CIP Classifications v1.0) requires the operator to declare the system's modality scope: which modalities it accepts as inputs, which modalities it produces as outputs, and which modalities were represented in its training data. These three are not necessarily the same — a text-to-image system accepts text but produces images and was trained on text-image pairs. The disclosure is operationally similar to the EU AI Act Article 53(1)(d) training-data summary requirement but extends beyond it to cover input and output modalities explicitly. The framework's Risk Gravity Model (Page 12f) implicitly addresses multimodal-media scenarios — the "Voice clone, brand client" and "Style appropriation, US output" scenarios both involve generative systems producing outputs across modalities — but did not previously name the category. From v2.8 onward, scenarios are explicitly classified by `CIP-Media-Class` to make the analysis tractable.
Rights held by performers, phonogram producers, and broadcasters in creative works, distinct from copyright. They include reproduction rights, making available rights, and communication rights. Neighbouring rights subsist independently of copyright and are not waived by AI ingestion.
Name, Image, Likeness, Publicity. Rights over the commercial use of a person's identity markers — name, face, voice, persona, likeness, and brand identity. NILP rights are engaged by voice cloning, deepfakes, AI-generated likenesses, and synthetic impersonation.
The licence type that governs what AI-generated content derived from a creative work may be used for. Output licence types include: Full Ownership, Shared/Split Rights, Attribution Required, Restricted Use, Non-Commercial Only, Platform-Controlled, and Blocked.
The documented history of where content comes from and how it has been transformed. In the CIP framework, provenance is captured in the Core Data Record and evidenced through the Rights Registry.
Rights granted by a state authority following a formal application, examination, and payment of fees. They do not exist until the registration is confirmed and are maintained only while renewal fees are paid. Examples include patents (granted after examination of novelty and inventive step), registered trade marks (granted after examination and opposition periods), and registered designs (protecting the visual appearance of a product). Registered rights are largely peripheral to the CIP framework, which focuses on subsisting rights — the automatic rights that arise at creation and persist through every stage of AI use, regardless of registration. The primary IP issues in AI pipelines are subsisting rather than registered. However, AI-generated outputs may engage registered trade marks where they reproduce a brand's distinctive signs, and legal practitioners should be aware of both categories.
The order and structure through which payments or royalty shares are allocated across rights holders following AI use of a creative work. The CIP framework provides for royalty automation through the Rights flow mechanism.
The rights-related information attached to a content record or asset within the framework — including the input licence class, transformation permissions, and output licence. Platform Certification requires 95% Rights Payload coverage across ingested content.
The verification system used to confirm rights-related status, credentials, and recognised outputs within the CIP framework. The Rights Registry supports CDR lookup, badge verification, and agency and platform listing.
Used to describe systems, actions, or participants that recognise and respond to the rights and consent conditions attached to creative content. A rights-aware platform identifies subsisting rights before ingesting content.
The rights that remain legally enforceable in a creative work, identity, performance, or dataset despite the asset having been used, processed, transformed, or reproduced through AI systems. The CIP framework — governed by Creative Intellectual Property Charity — exists to make these rights visible, evidenced, and enforced across the full AI content pipeline. Subsisting rights arise automatically at creation or qualifying act — no registration, application, or fee is required. This distinguishes them from registered rights (patents, trade marks, registered designs), which only come into existence after a formal state grant. The AI content pipeline almost exclusively engages subsisting rights: copyright, moral rights, neighbouring rights, NILP rights, database rights, and biometric rights all subsist from the moment of creation, and persist regardless of whether any registration was ever made.
Text and Data Mining opt-out. A declaration by a rights holder that their content may not be used for AI training or data mining. Under UK law (Data Use and Access Act 2025), EU law (AI Act Article 53), and US fair use analysis, opt-out mechanisms carry distinct legal weight.
A consent grant that is limited to a specific time period. After the period expires, the consent lapses and the rights holder must be approached again for any further use.
The payment or compensation owed to a rights holder whose creative work has been used as training data for an AI system. In the CIP framework, training markets are designed to pay per contribution path rather than per dataset inclusion.
A classification of the type of change applied to a creative work. Classes include: Original, Reproduction, Modification, Remix, Style Transfer, and Synthesis. Each transformation class has different rights implications depending on the input licence.