Versatile AI Risk Assessment

Glossary Systems and architecture

Fine-tuning

An existing model is trained further on your own data.

As of: September 2026 · Glossary with 28 terms

What is Fine-tuning?

In fine-tuning an existing model is trained further on your own examples to adapt style, domain language or task behaviour. The knowledge then sits in the weights, unlike RAG where it is supplied at runtime.

From a data protection view this is the trickier path. Personal data in the training examples cannot simply be deleted from a model again, and access and erasure requests become hard to satisfy. Clarify the legal basis, purpose limitation, and whether RAG achieves the same benefit without this risk.

Training dataGDPRRAG

Threats using this term (5)

Entries in the threat catalogue whose text uses the term. Every link leads to the full threat page.

Supply Chain – Datasets Training and fine-tuning data from external sources can be poisoned, flawed, or legally tainted. Training Data Poisoning Attackers inject manipulated or false data into the data an AI model learns from. Sleepy Agent (Time/Event-Triggered Hidden Instructions) Malicious logic lies dormant inside the model and only activates later: on a set date, at a specific event, or in a particular environment. Hate Speech and Discrimination The AI system produces content that demeans individuals or groups on the basis of protected characteristics such as origin, gender, … Misalignment The AI system does not pursue the goals its operator or users intend but whatever it was actually optimized for.

More terms from the subject area Systems and architecture.

GPAI and systemic risk Foundation models for many tasks, with their own duties and a separate tier for systemic risk. Agentic system (agentic AI) A system that uses a language model to carry out multi-step tasks on its own. Multi-agent system Several agents share a task and call each other.

From the term into the substance

The link leads to the place on the website where the term becomes practical; the overview shows every term in the glossary by subject area.

See data threats in the catalogue All terms in the glossary

Cite this term

For reports, policies or internal documents; the link leads directly to this term page.

“Fine-tuning”. Versatile AI Risk Assessment, glossary of AI risk analysis, as of September 2026.
https://www.versatile-ai-risk-assessment.com/en/wissensbasis/glossary/feinabstimmung/

← Back to the glossary