Adoption: The extent to which users effectively utilize and integrate a software-as-a-service solution into their daily operations and workflows.

‍AI assistant: A computer program utilizing artificial intelligence to perform tasks or services for a human user. These can range from scheduling appointments to answering questions and providing recommendations.

‍AI safety: The interdisciplinary study of how AI can be safely developed and used, focused on the prevention of accidents, misuse, or other negative consequences. A series of instructions for a computer program on how to analyze data in a certain way, such as recognizing patterns, so it can then learn from its analysis and accomplish tasks on its own.

‍Algorithm: A series of instructions for a computer program on how to analyze data in a certain way, such as recognizing patterns, so it can then learn from its analysis and accomplish tasks on its own.

‍Alignment: Refining an AI to better produce the desired outcome, and may refer to everything from moderating content to maintaining positive interactions with humans.

‍Application programming interface (API): Software code consisting of predefined commands, functions, and protocols that is used primarily to integrate one software application with another.

‍Artificial intelligence (AI): Computer systems capable of performing tasks that usually require human intelligence like visual perception, speech recognition, decision-making, and language translation.

‍Assisted authoring: Using technology to enhance or support the creation or editing of written content.

‍Audit trail: A chronological record of all changes and activities related to a contract.

‍Automated redlining: The delivery of AI-suggested redlined edits for existing clauses based on the guidance entered by the user, and usually occurring in the document editor of a CLM platform.B

Bias: Systematic and unfair disparities in the operation, outcomes, or decisions of AI systems. If unmitigated, AI trained on biased inputs can lead to unrepresentative outputs or even unfair or discriminatory outcomes against certain groups or individuals.

‍Big data: The large and complex sets of data generated by businesses, individuals, and devices. Big data requires special tools and techniques for effective analysis, processing, and storage.

‍Breach: The failure to fulfill any of the agreed-upon terms or conditions in a contract.

‍Business associate agreement (BAA): A contract between a covered entity and a business associate to protect personal health information (PHI) in accordance with HIPAA regulations.

ChatGPT: A chat-based assistant (sometimes referred to colloquially as a “chatbot”) launched by OpenAI that employs a large language model.

‍Clause: A section that specifies a particular condition, requirement, or term within the agreement.

‍Clause extraction: Automatic identification and collection of specific language from a contract to enable search, analysis, and reporting.

‍Clause library: A centralized repository of pre-approved legal clauses used to streamline and standardize the drafting of contracts.

‍Cognitive computing: A branch of artificial intelligence aspiring to create computer systems that can understand, reason, and learn like humans by using natural language processing, machine learning, knowledge representation and other techniques.

‍Contract AI: Applying artificial intelligence (AI) technologies to contract management and analytics. It involves using machine learning, natural language processing, and other AI techniques to assist in drafting, reviewing, and analyzing contracts.

‍Contract analytics: The use of AI and data analytics techniques to analyze contract data for various purposes, including, identifying trends and patterns, assessing risks, and evaluating global usage of fields and clauses.

‍Contract approval process: A predefined sequence of steps and approvals required for a contract to be authorized and executed. It helps ensure compliance with regulations, internal policies, and stakeholder requirements.

‍Contract approval matrix: A framework that outlines the authorization levels required for contract approvals based on certain criteria like contract value, type, or risk.

‍Contract authoring: The process of creating a contract document by defining the terms, conditions, and clauses that govern the agreement between parties. It involves drafting, reviewing, and editing the contract to ensure accuracy and compliance.

‍Contract automation: Using AI and other technologies to automate the creation, review, and management of contracts. It benefits an organization by streamlining the contract process, reducing errors, and improving efficiency.

‍Contract compliance: The adherence to the terms and obligations outlined in a contract throughout its lifecycle. It involves monitoring performance, enforcing terms, and resolving any discrepancies to ensure that both parties fulfill their obligations.

‍Contract execution: The final stage of the contract lifecycle where parties sign the agreement to make it legally binding. Execution involves obtaining signatures, exchanging copies of the contract, and fulfilling any conditions necessary for the contract to take effect.

‍Contract field: A specific data entry point within a contract document or CLM system for capturing and storing information, such as party names, dates, or contract terms.

‍Contract hierarchy: Sometimes referred to as “parent-child relationship”, contract hierarchy refers to the structured organization of contracts, showing their relationships and dependencies, such as master agreements and their subsequent amendments or schedules.

‍Contract intake: The initial process of capturing and entering contract details into the system for management and analysis.

‍Contract lifecycle management (CLM): The systematic and organized management of contracts from initiation through execution, renewal, and ultimately termination. CLM involves processes, strategies, and tools to streamline contract creation, negotiation, and compliance.

‍Contract monitoring and reporting: Tracking contract performance, analyzing data, and generating reports to assess compliance, financial impact, and risks associated with contracts. It helps stakeholders make informed decisions and optimize contract management processes.

‍Contract negotiation: The back-and-forth discussions and revisions (often colloquially referred to as “redlines”) between parties to reach mutual agreement on the terms of a contract. It involves clarifying expectations, resolving disputes, and finalizing the contract terms before execution.

‍Contract renewal: The process of extending or renegotiating an existing contract after its current term expires. It involves evaluating performance, reviewing terms, and negotiating any changes or extensions to the contract.

‍Contract repository: A centralized database or software system that stores all contracts and contract-related information in a structured manner. It serves as a secure location to access, manage, and track contracts throughout their lifecycle.

‍Contract template: A pre-designed format or structure that outlines the terms and conditions of a contract. Templates are used to standardize contract creation, ensure consistency, and facilitate the drafting process.

‍Contract termination: The conclusion of a contract's lifecycle when the agreement is legally ended or expired. Termination can occur due to completion of services, breach of contract, mutual agreement, or other specified conditions.

‍Contract workflow: A predefined sequence of processes and tasks for creating, reviewing, approving, and managing contracts.

‍Conversational AI: A form of AI enabling computers to understand, interpret, and respond to human language in a way that simulates human conversation. It is commonly used in chatbots, virtual assistants, and other types of applications that require natural language processing capabilities.

‍Counterparty paper: A contract or legal document drafted by the opposing party in a transaction or agreement.

Dashboard: A user interface that provides an overview of actionable insights related to a contract, a group of contracts, or the contracting process.

‍Data extraction: The act of identifying and retrieving information from typically unstructured or poorly structured sources for further processing or storage.

‍Data mining: The process of discovering patterns and relationships in large datasets using statistical and machine learning algorithms.

‍Deep learning: A subfield of machine learning involving the building of neural networks with multiple layers that can learn increasingly complex representations of data.

‍Document intelligence: The capability of a computer system to understand and extract information such as text, images, and tables from documents. This can include document analysis, document recognition, document summarization, and document generation.

‍Document management: The process of organizing, storing, and retrieving electronic or paper documents.

‍Document recognition: The process of automatically identifying and extracting data from a document, such as text or images.

E-Signature: A digital form of signing documents electronically.

Feedback loop: A process where the output of a system is used as input for the same system, creating a circular chain of cause-and-effect and influencing future output. Feedback loops can be positive or negative, depending on whether they amplify or dampen the effects of the input. These are used in a wide range of fields, from computing, engineering and economics to biology and psychology, to help explain and model complex systems.

‍Field: A specific data entry point within a contract document or CLM system for capturing and storing information, such as party names, dates, or contract terms.

‍Force majeure clause: A contract provision that frees both parties from obligation if an extraordinary event, such as a natural disaster, prevents performance.

‍Foundation model: A very large LLM or AI system pre-trained on an immense quantity of data, and can therefore be applied to a wide array of tasks..

GAN (Generative Adversarial Network): A type of neural network that consists of a generator and a discriminator which are trained together in a competitive game-like setting to generate new data that's similar to a given training set.

‍Generative AI: A form of artificial intelligence that has the ability to generate original content like music, images, videos, and text by using complex algorithms. Unlike other forms of AI that are mainly used for classification or prediction, generative AI is used commonly in creative fields like art, music, and design, as well as in industries like marketing and advertising, and is now being applied to CLM for the legal sector.

‍Governing law clause: A contract clause that determines the state or country's laws that will be used to interpret the contract.

‍GPT: A "generative pre-trained transformer" AI language model that powers chat-based assistants like ChatGPT, among others. It has been pre-trained on a large pool of text data and can generate human-like responses to a wide range of prompts.

‍Guardrails: An essential facet of capable AI solutions ensuing they have built-in protections to help the user verify the correctness of AI output and understand how the AI has generated that output.

Hallucination: When an AI generates output that is not accurate or relevant to the input data. This can happen when the AI model is trained on biased or incomplete data, or not properly calibrated to recognize and account for outliers or unusual patterns in the data. This is a common challenge in the development of AI systems and is mitigated by improving data quality, developing more robust training algorithms, and better error detection and correction mechanisms.

Indemnification clause: A contract provision where one party agrees to compensate the other for certain losses.

‍Intelligent contract analysis: This is the use of AI algorithms to analyze contracts and extract key data and insights. It helps in identifying potential risks, ensuring compliance, and providing actionable insights for decision-making.

‍Intelligent document generation: Refers to the use of AI to automatically generate contracts and legal documents based on predefined templates and rules. It helps to save time and reduces the chances of errors in contract drafting.

‍ISDA master agreement: A standardized contract published by the International Swaps and Derivatives Association to govern the terms of derivatives transactions between parties.

Knowledge graph: A type of graph database that represents knowledge as entities (nodes) connected by relationships (edges). In contract AI, knowledge graphs can be used to represent and structure contract data, making it easier to extract insights and analyze contracts.

Large language models (LLMs): AI models that are trained on expansive amounts of data, such as a significant portion of the public internet or a large organization’s database and documentation, and that may contain billions of parameters.

‍Legal operations: Team that manages the activities and processes that optimize the legal department's efficiency and effectiveness in a business.

‍Limitation of liability clause: A contract clause that limits the amount one party can be liable to the other.

‍Long short-term memory: A recurrent neural network designed for overcoming vanishing gradients by using a gated memory cell.

M&A due diligence: The comprehensive appraisal of a business by a prospective buyer to evaluate its commercial, financial, and legal circumstances before a merger or acquisition.

‍Machine learning (ML): A branch of artificial intelligence that enables machines to learn from past experiences and data, without the need for explicit programming. ML algorithms are used in contract AI to analyze and classify contracts, identify fields and clauses, and extract key information.

‍Master service agreement: A contract that outlines the general terms and conditions between a service provider and a client for ongoing services.

‍Model: In AI, this is a program that employs learning algorithms to generate conclusions, predictions, or actions by analyzing large datasets, usually with an ability to improve its accuracy over time through machine learning or deep learning.

Natural language processing (NLP): A branch of AI focusing on the interaction between computers and human language. In contract AI, NLP extracts information from contracts after processing and analyzing the natural language data within them.

‍Neural network: A set of algorithms designed to recognize patterns and relationships in data by simulating the structure and function of the human brain.

‍Non-disclosure agreement (NDA): A legal contract to keep certain information confidential, with mutual NDAs binding both parties to secrecy, while unilateral NDAs bind only one party.

OpenAI: An AI research lab founded in 2015 that achieved mainstream adoption of generative AI in 2022 with its release of  ChatGPT, an AI-powered chat-based assistant.

‍Optical character recognition (OCR): Also known as optical character reader, this is the electronic or mechanical conversion of images of typed, handwritten or printed text into machine-encoded text. This can be from a scanned document, a photo of a document, subtitles within an image, or other sources. The use of OCR is key to a contract AI system being able to “ingest” paper documents or PDFs.

‍Order form: A contract document that specifies the products or services being purchased, including quantities, prices, and terms.

Parameter: A variable an AI model learns during training that it then employs to make predictions or decisions.

‍Post-merger integration: The process of combining and restructuring two companies into one entity after a merger or acquisition.

‍Predictive contract analytics: Utilizing AI algorithms and historical contract data to predict outcomes and trends related to contract drafting and negotiation, helping in optimizing contract terms and making data-driven decisions.

Recurrent neural network: A neural network that can process sequential data by maintaining an internal memory state, which allows it to learn long-term dependencies.

‍Reinforcement learning: A kind of machine learning where an AI learns to make decisions by trial and error, receiving feedback in the form of rewards or penalties.

‍Responsible AI: The development and use of artificial intelligence in a way that is ethical, transparent, and accountable, meaning AI systems should be designed to respect human rights, protect privacy, and avoid bias and discrimination. Responsible AI also involves ensuring AI is used in a way that benefits society as a whole, rather than just a particular group or individual. Additionally, it involves creating systems that are secure, reliable, and safe, and that can be audited and explained to users.

‍Retrieval augmented generation (RAG): Combining generative and retrieval-based approaches to produce high-quality responses to user queries by using a retrieval component to identify relevant information then used to generate a response. This approach enables RAG models to provide more accurate and contextually relevant responses to user queries and has been applied in virtual assistants, chatbots, and question-answering systems.

Semantic analysis: The process of comprehending the meaning of text or language based on its context. In contract AI, semantic analysis is used to extract and classify contract clauses, identify relationships between clauses, and derive insights from contract data.

‍Smart contracts: These are self-executing contracts where the terms of an agreement are written directly into the code. Not directly related to contract AI, smart contracts leverage blockchain and cryptographic technologies to automate contract execution and enhance security.

‍Structured data: Information organized and formatted in a precise way, as in a table or database, where it follows a defined data model or schema.

Temperature: A parameter influencing LLM output, determining whether it is random and creative or predictable. A higher temperature will result in lower probability and therefore more creative outputs.

‍Termination clause: A clause that outlines the conditions under which the contract can be terminated.

‍Third-party paper: A contract or legal document drafted by the opposing party in a transaction or agreement.

‍Training data: The dataset that is used to train an AI system.

‍Transfer learning: A technique that involves reusing a pre-trained model for a similar task, which can often result in improved performance and faster training.

Unstructured data: Information that does not follow a defined format or structure; examples include reports, articles, and emails.

‍Unsupervised learning: A process where AI is given data or information that has not been labeled or structured by humans, so the AI must discover patterns without human guidance.

Vendor agreement: A contract that outlines the terms and conditions of a business arrangement between a buyer and a supplier of goods or services.

Workflow: A predefined sequence of processes and tasks for creating, reviewing, approving, and managing contracts.