Artificial intelligence, LLMs, generative AI, prompt engineering. The list goes on.
You don’t need to follow every AI headline to see that this language is quickly becoming part of the conversation. And when you’re busy practicing law, deciphering tech-speak isn’t high on your to-do list.
That said, understanding a few key terms can help you evaluate tools, ask sharper questions, and make informed decisions about how AI could fit into your firm.
Consider this your plain-English guide to the AI terms you’re most likely to encounter in emails, articles, webinars, product demos, and conversations about legal technology.
Smokeball AI Glossary
Algorithm: A set of instructions a computer follows to complete a task or solve a problem.
Algorithmic Bias: When an AI system produces unfair or skewed results because of the information it was trained on.
Artificial Intelligence (AI): Technology that helps computers perform tasks that normally require human thinking, such as understanding language, recognising patterns, making recommendations, or generating content.
Automation agent: Technology that performs repetitive tasks automatically, reducing the need for manual work.
Chatbot: A tool that uses AI to have conversations and answer questions.
ChatGPT: An AI assistant created by OpenAI that can answer questions, draft content, summarise information, and more.
Claude: An AI assistant created by Anthropic that can answer questions, draft content, summarise information, and more.
Cybersecurity: The safeguards used to protect systems, data, and information from unauthorised access, breaches, or cyberattacks.
Data Privacy: How personal information is collected, stored, shared, andprotected. For law firms, this includes considering confidentiality obligations, privacy requirements and whether client or personal information may be handled or stored overseas.
Deepfake: An AI-generated image, video, or audio recording that appears real but has been manipulated or completely created by AI.
Encryption: A security measure that helps protect data from unauthorised access.
Ethics: The professional responsibilities and considerations involved in using AI, including issues such as confidentiality, competence, supervision, and accuracy.
Fine-Tuning: Additional training given to an AI model to make it better at a specific task, industry, or type of work.
Foundation Model: A large AI model trained on broad information that serves as the foundation for tools like ChatGPT, Claude, and other AI applications.
Generative AI: AI that creates new content, such as text, images, summaries,audio, or ideas.
Hallucination: When AI confidently provides information that is incorrect, madeup, or misleading.
Large Language Model (LLM): The technology behind tools like ChatGPT. LLMs are trained on large amounts of text and can understand and generate human language.
Machine Learning (ML): A type of AI that learns patterns from data instead of being programmed with specific rules.
Model: TheAI system itself. A model is what has been trained to recognise patterns and generate responses.
Natural Language Processing (NLP): Technology that helps computers understand and respond to human language.
Open Source: Software whose underlying code is publicly available for anyone to review, modify, or improve (often where data privacy is of higher concern)
Predictive AI: AI that uses patterns to make recommendations or predictions about what might happen next.
Predictive Analytics: Technology that uses historical data and AI to help forecast future outcomes or trends.
Prompt: The question, instruction, or request you give an AI tool.
Prompt Engineering: The practice of refining prompts to get better, more accurate, or more useful responses from AI.
Responsible AI: The practice of developing and using AI in a way that is ethical, secure, transparent, and trustworthy.
Semantic Search: Search technology that focuses on meaning and intent not just keywords to find more relevant information.
Token: A small piece of text AI uses to process language. Think of tokens as the building blocks AI reads and writes.
Training Data: The information AI learns from before it can answer questions or complete tasks.
Transparency: The ability to understand how an AI system reached a recommendation, conclusion, or response.
You know the lingo, let’s get to work
You don’t need to memorise every AI term to benefit from it. Knowing the basics simply makes it easier to understand the tools available to you and how they can support your work.
