Since 2 August 2026, people must be told when they're dealing with AI. Stricter rules for AI used in hiring and credit decisions follow on 2 December 2027. Find out where your business stands in two minutes.
Where does your AI reach?
A quick indicator, not a legal assessment. Only a full audit gives you a documented classification.
If an AI tool talks to your customers, creates content they see, or reads emotions, people have to be told. For most businesses this means clear notices and a few settings changed. It's quick and inexpensive, and most businesses haven't started.
What Article 50 asks of you →AI used to screen job applicants, assess credit or price insurance faces the full weight of the Act: human oversight, records, and informing the people affected. That takes months to put in place properly, so businesses that start now avoid a scramble in late 2027.
See the full timeline →We don't build AI and we don't sell software. We make the AI you already use defensible, and we stay on as your tools and the rules change.
Including features switched on inside software you already pay for, which is where most businesses miss things.
Banned, high-risk, transparency duty, or minimal. We set out whether you're the provider or the deployer and which duties follow.
Transparency notices, an AI register, a risk register and a prioritised plan, written in plain English.
A short monthly check-in covers new tools you've adopted, rule changes and what they mean for you.
Maximum fine for using a banned AI practice.
Art. 99(3)Maximum fine for breaching most other duties, including transparency and high-risk rules.
Art. 99(4)For SMEs and start-ups, each cap is whichever of the fixed amount or the percentage of turnover is lower.
Art. 99(6)You deal directly with the people doing the work, from the first call to the monthly check-in.
About us →Yes. It is an EU Regulation, so it applies directly in Ireland without needing to be turned into Irish law first. It covers businesses that build AI and businesses that use it.
If your business uses AI that affects customers, staff or job applicants, some of the Act's duties probably apply to you already.
Partly. The Act splits duties between the provider who builds a tool and the deployer who uses it. Vendors carry a lot of the obligations, but deployers have their own: telling people when they're dealing with AI in some cases, and, for high-risk uses such as hiring, human oversight, keeping logs and informing the people affected.
Working out which role you're in for each tool is one of the first things an audit settles.
The Act gives smaller businesses simplified templates. A template tells you what to fill in, not what applies to you. The real work is classifying each AI tool correctly and being able to defend that classification. A template filled in wrongly is still wrong.
A fixed fee, agreed before we start, based on how many AI tools you use. There's no hourly billing and no open-ended engagement. Book a call and we'll quote you.
No. We're a compliance service. We inventory your AI tools, classify them against the Act, produce the documentation and keep it current. We don't give legal advice, and we'll tell you if a question needs a lawyer.
Up to €35 million or 7% of worldwide turnover for banned practices, and up to €15 million or 3% for breaching most other duties, including the transparency and high-risk rules. For SMEs the cap is whichever of the two amounts is lower (Article 99).
Five questions, two minutes, and a written result in your inbox.