National newscaster Charity Digital explores what The European Union’s Artificial Intelligence Act (EU AI Act) is and what it could mean for charities in the UK.
The
EU AI Act is the world’s first law that
regulates artificial intelligence (AI). The law creates strict and comprehensive rules for AI developers and "deployers" (or people who make use of the technology).
It has been
designed to protect you by placing AI into four risk categories:
Minimal-risk,
Limited-risk,
High- risk, and
Banned.
Because of its broad reach, the law affects you even if you live in the UK, provided the AI system operates in the EU or
affects your daily life.
MVA offers a quick rundown of the categories and how you may be impacted:
Minimal-Risk AI
Everyday AI, like basic video game opponents or spam filters come into this category; these systems are currently unregulated and so
carry no special rules.
Limited-Risk AI
Systems like
chatbots and
deepfakes fall into this category:
Creators of these systems must tell you when you are interacting with AI; as an example, a chatbot must explicitly notify you that it is not a human, and that manipulated audio or video (known as deepfakes) must be labelled.
This is
to make sure you are not tricked or wrongly influenced by the given system you are using.
High-Risk AI
AI used in critical areas is heavily restricted but still allowed if it passes strict safety and bias tests.
If you apply for a job, a bank loan, or an education program, the AI used to score or evaluate your application must be
heavily monitored by real humans to prevent discrimination.
And as a user, you have the right to request clear explanations for these decisions.
Banned AI (Unacceptable Risk)
The law completely
prohibits AI systems that threaten your safety, rights, or freedoms.
As an example, you are protected from "Social Scoring" - where a government or company ranks you based on behaviour - manipulating your mind or habits using subliminal tricks, and
untargeted scraping of internet or CCTV photos to create facial recognition databases.
How it can Impact UK Charities
The EU AI Act is the world’s first comprehensive AI law. It sorts artificial intelligence (AI) by its risk level to people’s rights and safety; and if a UK charity operates in Europe or uses AI systems that impact people in the EU, the law would apply to them
Even though the UK is not in the EU, the
law has a global reach, and charities might need to follow these rules if they:
- Target the EU: offer services, campaigns, or fundraising to people in the EU.
- Uses third-party AI: use AI tools (like translation or donor-matching software) that process data from EU residents.
- Build custom AI: develop AI models and deploys them in the EU market.
For high-risk systems, charities face heavy technical paperwork and oversight, and for limited-risk tools (like AI-written newsletters or fundraising chatbots), charities must make sure they
clearly disclose that content is AI-generated.
Charities that break the rules can face huge fines, like the regulations set for
GDPR.
Newscaster
Charity Digital expand on this subject and more in their feature about the EU AI Act.
From
How AI is regulated within the UK To
The broader impact of the EU AI Act, they explain what the European Union’s Artificial Intelligence Act is and what it could mean for UK charities.
To
read the feature, visit the Charity Digital
website.