![](https://www.gordonsllp.com/wp-content/uploads/2024/05/AI-2-1024x535.jpg 1024w, https://www.gordonsllp.com/wp-content/uploads/2024/05/AI-2-300x157.jpg 300w, https://www.gordonsllp.com/wp-content/uploads/2024/05/AI-2-768x401.jpg 768w, https://www.gordonsllp.com/wp-content/uploads/2024/05/AI-2.jpg 1200w)
Decoding AI: Key terms in AI agreements
Thursday 20th June 2024
As AI services become increasingly integral to business operations, it’s important to grasp the industry’s key terms. This knowledge helps differentiate various systems and evaluate their suitability. While AI agreements and statutes might offer their own definitions, this primer is a foundational guide for those new to the AI lexicon.
Adaptivity: The ability to see patterns and make decisions in ways not directly envisioned by human programmers.
Algorithm: A sophisticated programming sequence that empowers an AI system to learn and function.
Alignment: The process of ensuring an AI system’s goals and behaviours are in line with human values and intentions.
Artificial Intelligence (AI): Software or systems capable of autonomously executing complex tasks through human-like cognitive processes.
AI agents: Autonomous AI systems that perform multiple sequential steps – including browsing the internet, sending emails, or sending instructions to physical equipment – to try and complete a high-level task or goal.
AI deployers: Any individual or organisation that supplies or uses an AI application to provide a product or service to an end user.
AI developers: Organisations or individuals who design, build, train, adapt, or combine AI models and applications.
AI end user: Any intended or actual individual or organisation that uses or consumes an AI-based product or service as it is deployed.
AI Output: The novel data produced by an AI system in response to given instructions.
Application Programming Interface (API): A set of rules and protocols that enables integration and communication between AI systems and other software applications.
Autonomous: Capable of operating, taking actions, or making decisions without a human’s intent or oversight.
Capabilities: The range of tasks or functions an AI system can perform and the proficiency with which it can perform them.
ChatGPT: An abbreviation for Chat Generative Pre-Trained Transformer, this AI chatbot leverages natural language processing to craft responses.
Conversational AI: AI systems designed to mimic human conversation, functioning as chatbots or virtual assistants.
Data Governance: Policies and processes that govern the accessibility, usability, integrity, and security of the data used by the AI system.
Data Mining: The technique of extracting valuable insights from vast datasets, utilising machine learning algorithms for analysis.
Deep Learning: A machine learning approach that mirrors human cognitive learning, employing intricate neural networks to discern patterns within extensive datasets.
Extractive AI: AI models that specialise in distilling relevant information from supplied data.
Foundation models: Machine learning models trained on very large amounts of data that can be adapted to a wide range of tasks.
Generative AI: AI models that can create fresh data (such as text, images, or videos) from existing information.
Interoperability: The ability of the AI system to work with other systems or software, which can be crucial for integration into existing technology environments.
Machine Learning (ML): A learning process akin to human experience, where AI systems apply provided training data to create algorithms and methods for complex task execution.
Model Drift: The change in model performance over time due to evolving data patterns on which the model was not trained initially.
Natural Language Processing (NLP): A machine learning variant enabling AI systems to interpret and process human language.
Neural Network: A series of algorithms that functions analogously to human brain connections, facilitating learning and decision-making in AI systems.
Service Level Agreements (SLAs): Formal agreements that define the level of service expected from the AI system provider, including uptime, performance, and support parameters.
Performance Metrics: Criteria used to evaluate the AI system’s effectiveness and efficiency in achieving its intended tasks.
Prompt: The user’s message to an AI system, outlining the desired task.
If you have any questions about adopting or licensing an AI system or need advice on your AI agreements, please get in touch with one of our AI and technology law experts.
This article is part three in a series on AI.