Resources

Glossary

Key terms and definitions for enterprise AI in regulated environments.

AI: Stands for artificial intelligence, which is the simulation of human intelligence processes by machines or computer systems. AI can mimic human capabilities such as communication, learning, and decision-making.
Cloud Computing: The delivery of computing services — including servers, storage, databases, networking, software, analytics, and intelligence — over the internet ("the cloud") to offer faster innovation, flexible resources, and economies of scale.
Custom AI: Is the process of developing a company-specific AI solution targeting a particular problem. Since custom AI software is developed for a single business, it needs to satisfy the business's specifications and expectations.
Deep Learning: A function of AI that imitates the human brain by learning from how it structures and processes information to make decisions. Instead of relying on an algorithm that can only perform one specific task, this subset of machine learning can learn from unstructured data without supervision.
Deterministic AI: AI whose decision path returns the same output for the same inputs, model versions, and configuration, so a result can be re-run and checked.
Ensemble AI: Is the process by which multiple models, such as classifiers or experts, are strategically generated and combined to solve a particular computational intelligence problem.
Generative AI: Models that generate new text, images, code, or other content from patterns in their training data. Typical transformer-based generative models can produce different outputs for the same input. Frontier Foundry's generative AI is custom and not transformer-based: it is deterministic, directly auditable, and never the final decision-maker.
Hallucination: When an artificial intelligence model detects non-existent patterns or incorrectly analyzes data, leading to inaccurate or misleading outputs.
Large Language Models (LLMs): Type of language model notable for its ability to achieve general-purpose language understanding and generation.
Machine Learning (ML): A subset of AI that incorporates aspects of computer science, mathematics, and coding. Machine learning focuses on developing algorithms and models that help machines learn from data and predict trends and behaviors, without human assistance.
Model Risk Management: The governance, testing, and monitoring a bank uses to control the risk of decisions based on model output. In the U.S., the agencies set out their principles in SR 26-2.
Natural Language Processing (NLP): A type of AI that enables computers to understand spoken and written human language. NLP enables features like text and speech recognition on devices.
Neural Networks: A deep learning technique designed to resemble the human brain's structure. Neural networks require large data sets to perform calculations and create outputs, which enables features like speech and vision recognition.
Post-Quantum Cryptography (PQC): Cryptography designed to resist attacks from quantum computers, including the NIST standards ML-KEM, ML-DSA, and SLH-DSA (FIPS 203, 204, and 205).
Quantum Computing: Computing that uses quantum-mechanical effects such as superposition and entanglement to solve certain classes of problems — for example, factoring the large numbers that underpin today's public-key encryption — that are impractical for classical computers. This is why Frontier Foundry ships post-quantum cryptography (NIST FIPS 203, 204, and 205) as standard.
SR 26-2: Revised U.S. banking agency guidance on model risk management, issued in April 2026, which replaced SR 11-7. It is addressed mainly to banking organizations with over $30 billion in assets. Its definition of a model excludes deterministic rule-based processes with no statistical, economic, or financial theory behind them, and generative and agentic AI are outside its scope.