Impact Assessment
Last Updated: July 29, 2026 | By Mihail Sebastian | AI Dictionary
An evaluation of how deploying an AI system will affect people and society, from privacy and fundamental rights to jobs and the environment.
What is an Impact Assessment?
An impact assessment evaluates how deploying an AI system will affect the people and society around it: privacy, fundamental rights, employment, the environment. Where most testing asks whether the system works, an impact assessment asks what the system does to the world when it works exactly as designed. Several laws now require one before deployment.
How an Impact Assessment Works
The assessment starts by identifying everyone the system touches: not just its users, but the people it makes decisions about and third parties downstream. For each group, the assessor maps plausible effects (loss of privacy, discriminatory outcomes, job displacement, the energy cost of compute), rates their severity and likelihood, and records the measures that will reduce them.
The result is a document that decision-makers sign off on before the system goes live, and revisit when it changes.
Two legal instruments anchor the practice in Europe. Under the GDPR, a Data Protection Impact Assessment (DPIA) is required when data processing is likely to create a high risk to people’s rights and freedoms, which covers much of the AI that profiles individuals.
The EU AI Act adds a Fundamental Rights Impact Assessment (FRIA) for certain deployers of high-risk systems, including public bodies.
Impact Assessment vs AI Assessment
The practical difference: an impact assessment looks outward at a system’s effects on people and society, while an AI assessment looks inward at the system itself, its accuracy, robustness, and risks.
The two overlap on fairness, since a biased model is both a technical defect and a social harm, but they answer different questions.
A system can pass every technical assessment and still fail an impact assessment because of what it does to the people subject to its decisions.
Example of an Impact Assessment
The Dutch childcare benefits scandal shows what an impact assessment exists to prevent. The Dutch tax authority used a risk-scoring system to flag childcare benefit applications as potentially fraudulent, and holding a second nationality was among the inputs that raised a family’s risk score.
Tens of thousands of families were wrongly accused of fraud and ordered to repay benefits; the fallout led the Dutch government to resign in January 2021.
An impact assessment run before deployment asks exactly the questions this system failed: who does the scoring target, is nationality a legitimate input, what happens to a family that is wrongly flagged, and how do they contest the decision.
Related AI terms: AI Assessment · AI Audit · AI Risk · Risk Management · EU AI Act
Did you like the Impact Assessment gist?
Learn about 250+ need-to-know artificial intelligence terms in the AI Dictionary.
Mihail Sebastian — Writes about AI governance, regulation, and the technology behind them. Placeholder bio — replace with a real credential line. About