8xav��������

Artificial Intelligence (AI) is a field within computer science that is dedicated to building machines that can perform tasks requiring human intelligence. These tasks include pattern recognition, understanding language, problem solving, learning from data and supporting decision making.
AI is an umbrella term encompassing a wide range of fields, including machine learning, natural language processing, computer vision and robotics. AI is used today in many parts of our daily lives, including recommendation systems, virtual assistants, medical diagnostics, scientific discovery, and autonomous vehicles.
Researchers at the 8xav�������� of Luxembourg work on the basic principles of AI and on its applications in health, finance, law, cyber security, social sciences and environmental research.

• Learn more about our research

These terms are often used interchangeably, but they are different ideas.

Artificial intelligence (or AI)

Artificial intelligence (or AI) is the broad science of designing systems that can do intelligent tasks.

Machine learning (ML)

Machine learning (ML) is a subset of AI that allows algorithms to learn patterns from data, rather than being explicitly programmed for each task.

Generative AI

Generative AI is a type of AI that is able to generate new content, such as text, images, software code, music, or videos. Examples include chatbots, image generators and AI coding assistants.

In summary:

• AI is the larger field.
• Machine learning is one way of building AI systems.
• Generative AI is a kind of AI that is all about creating content.

Large Language Models (LLMs) are AI systems trained on vast amounts of text to understand and generate human language.
They can answer questions, summarise documents, translate languages, write code and help with creative tasks. Famous examples include ChatGPT, Microsoft Copilot, and other conversational AI systems.
Large language models can produce impressive outputs, but they do not really “understand” information like humans do. They did not create words and patterns, but predicted likely words and patterns from their training data. This means they can sometimes produce factually inaccurate or misleading information, so human oversight is always necessary.

Many AI systems learn through a process called “machine learning.”
During training, algorithms analyse vast datasets to find patterns and relationships. The system improves its performance in specific tasks by processing examples many times. For example, AI that is trained to recognise tumours in medical images is trained on thousands of annotated scans. It gets better at spotting visual patterns tied to disease over time.
The quality, variety and accuracy of the training data are critical to the performance of an AI system.

AI systems can make mistakes, produce biased outputs or wrong information.
Any of these problems can occur for any of the following reasons:

• inadequate or biased training data
• flaws in the underlying algorithms
• lack of context
• dynamic real-world conditions.

Generative AI systems can also produce answers that sound plausible but are wrong, a phenomenon called “hallucination.”
Hence, explainability, transparency, fairness and accountability are gaining increasing importance in the research community to ensure that AI systems are used responsibly and safely.
At Uni.lu, the ULIDE focuses on ethical issues related to potential risks of AI, including the displacement of jobs, privacy concerns, and the possibility of AI systems making harmful decisions.

• More about this subject on the Institute for Digital Ethics (ULIDE) page

Yes. Europe is among the first regions in the world to develop a comprehensive legal framework for AI.
The EU’s AI Act offers risk-based rules for the creation and implementation of AI systems. Its purpose is to foster innovation while safeguarding fundamental rights, safety and trust.
Researchers from the 8xav�������� of Luxembourg are taking part in discussions on AI governance, regulation, ethics and trustworthy AI, helping to ensure the responsible development of these technologies.

AI is already supporting many applications in different sectors.

AI in Healthcare

Healthcare AI can help analyse medical images, assist in diagnosis, discover new drugs and enhance personalised treatment

AI in Finance

AI is used in fraud detection, risk analysis, financial forecasting and customer support.

AI in Cybersecurity

AI can detect suspicious behaviour, identify cyberattacks and help organisations respond faster to emerging threat.

Researchers at the 8xav�������� of Luxembourg explore AI applications in these and many other domains, frequently i close collaboration with partners from industry and the public sector.

• Discover our areas of expertise in AI

As AI systems become more powerful, it becomes more and more important to understand how they make decisions.

Explainable AI is the effort to make AI systems more transparent and understandable to its users.
Trustworthy AI goes beyond simply ensuring reliable systems; it aims to make them fair, secure, accountable and aligned with human values.
These are especially important principles in sensitive areas such as healthcare, justice, finance and public services, where AI-assisted decisions can have major consequences.

Artificial Intelligence is acknowledged as a main driver of innovation, economic growth and scientific progress.
Luxembourg wants to become a leader in human-centric and trustworthy AI through investment in research, infrastructure, education and cooperation between academia, industry and government.
The country’s strategy is to develop AI technologies that provide societal benefits while upholding high standards of ethics, transparency and trust.