Home AI Compass Builder
Fine-tuning
Further training of an existing model on your examples, to change its style or teach it a narrow task.
Builder path · 5 of 6
What it is
Fine-tuning adjusts a model's weights using a set of examples. It is good for a consistent format, tone or a specialized task. It is usually not the best way to give a model current facts — RAG does that better.
Example
A model fine-tuned on labeled support tickets sorts new tickets into the right category.
What can go wrong
Training data can be memorized and later repeated; a fine-tuned model inherits the errors and biases of its examples.
How to secure it
Clean and minimize the training data, check your right to use it, and evaluate the tuned model before and after deployment.
EU rules
Substantially modifying a model or an AI system can change your role, and your obligations, under the EU AI Act — check before you start.
Plain-language summary, not legal advice.
Related
Related service: Applied AI