Introduction to AI Guide with a focus on Counter Fraud - Glossary of Terms - Key terms and meanings

Nov 2, 2026

Uploaded by Oliver

Guide introducing artificial intelligence with an emphasis on counter fraud. Defines core concepts including generative AI, large language models, voice cloning, deepfakes, ethical AI, and machine learning, supported by a glossary and timelines of AI milestones. Explains how fraudsters adopt increasingly sophisticated data-driven techniques to exploit organisational vulnerabilities. Outlines why public-sector teams must keep pace, understand potential fraud threats, and apply ethical principles and data management to reduce strategic, operational, and reputational risks.

LanguageEnglish

CategoryTechnology

FormatPDF

Pages21

Reading Time53 min

AccessFree to download

Introduction

Glossary of Terms

The Rise of Artificial Intelligence

What is Artificial Intelligence (AI)?

Supervised vs Unsupervised learning

Introduction to AI Guide with a focus on Counter Fraud Glossary of Terms Key terms and meanings 1 Artificial Intelligence (AI) Machine driven capability to achieve a goal by performing cognitive tasks. | Generative AI | AI systems that create new content e.g. ChatGPT, generate text and images from text prompts-some use images to create audio and video content. | | --- | --- | | Large Language Models | Models trained on large volumes of text based data-typically from the internet. | | Voice Cloning | Use of AI technology to create a simulation of a person’s voice. | | Deep fake | Videos or images that use a form of AI to digitally manipulate existing content e.g. replacing images of faces with someone’s likeness. Deep fake can also be known as synthetic media | | Ethical AI | Used to indicate the development, deployment and use of AI that ensures compliance with ethical norms, including fundamental rights as special moral entitlements, ethical principles and related core values. It is the second of the three core elements necessary for achieving Trustworthy AI. | Machine Learning The use of algorithms that find patterns in data without explicit instruction. A system might learn how to associate features of inputs such as images with outputs such as labels. 1 Partly sourced NCA, GenAI Threat Assessment 2023 Introduction Technology does not stand still, whether we consider this from a counter fraud perspective or the view of the fraudsters we face. When technology evolves it can be harnessed by fraud practitioners to great benefit, but equally criminals and fraudsters will work at pace to embed these advances into their toolkits to attack systems and processes. Fraudsters are able to use increasingly sophisticated methods, relying on the systematic analysis of large amounts of data inan effort to identify and exploit vulnerabilities that might exist in our organisations for their own gain. Letting fraudsters lead the way in the use of Artificial Intelligence (AI) technology is not an option-so it is our collective role and responsibility in counter fraud to keep pace with developments and understand the impact and potential fraud threats they may bring and understand the opportunities that may arise. This short guide introduces, and we hope demystifies, AI, and signposts you onwards to build your knowledge and awareness. Fraudsters are able to use increasingly sophisticated methods to identify and exploit vulnerability The rise of AI Artificial intelligence is not new but we have seen accelerated coverage in the media and as a hot topic at public and private sector events in recent years. This is because access to AI tools has become more commonplace. GenAI platforms like ChatGPT are now widely available and used by the public in a variety of ways. The rise of Artificial Intelligence does present a huge opportunity for those working in the public sector to detect and prevent fraud, at pace, using large quantities of information data. This alignsand supports the modern fraud approach which focuses on a deep understanding of risk and the use of data and intelligence to find fraud and irregular payments. When using data and AI it is important that users consider potential strategic, operational and reputational risks that may arise if key principles, ethical considerations and data management processes are not adhered to. A brief timeline of AI2 Alan Turing-Computer machinery and intelligence- ‘Turing Test’ ‘Eliza’-first chatbot by Joseph Wiezenbaum Deep Blue IBM chess computer wins v’s world champion Garry Kasparov Dragon software introduced ‘Alexa’ virtual assistant by Amazon, learning from queries ‘Siri’ intelligence speech assistant introduced in i-phone 4 GPT-3 is released-enabling automated conversations 2022 Chat GPT released -AI explodes! 1950 1956 Arthur Samuel IBM 701 invent checkers game ‘machine learning’John McCarthy ‘the facts of AI’and Marvin Minsky logical

FAQs

What does the guide cover about artificial intelligence in counter fraud?

It introduces AI and demystifies key concepts while linking developments to fraud threats and opportunities for fraud prevention.

What are generative AI, large language models, and deepfakes?

Generative AI creates new content from prompts; large language models are trained on large text datasets; deepfakes digitally manipulate existing media using AI.

How does supervised learning differ from unsupervised learning?

Supervised learning trains on labelled datasets to classify or predict outcomes, while unsupervised learning uses unlabelled data to discover patterns without explicit guidance.

Nov 2, 2026

Uploaded by Oliver