The EU started enforcing the AI Act. Three things US firms should copy now.

On August 2, 2026, the European Union began enforcing its AI Act. If you run a firm in the United States, that law does not bind you. But three of its rules are simple, cheap, and worth copying right now, because they build trust with clients and put you ahead of where US expectations are heading. Aday Interactive, Inc. builds AI systems for regulated firms, and these are the three moves we would make first.

The short version

US firms are not bound by the AI Act. But three of its moves are cheap, fast, and good for business no matter where you operate: tell people when they are talking to AI, label the content your AI makes, and keep evidence of how your AI works. Each one builds client trust today and puts you ahead of where US rules are clearly heading. Architect for trust now. It is far cheaper than retrofitting it after a client, a board, or a regulator asks the question first.

Here is what actually changed. As of August 2, the EU AI Office and national authorities can enforce the rules for general-purpose AI models, the ban on prohibited practices, and a set of new transparency obligations. The most visible one: interactive AI must tell users they are dealing with AI, deepfakes must be labeled, and AI-generated content must carry machine-readable provenance marks. The heavier obligations for high-risk AI systems were postponed to December 2027 and August 2028, so what is live today is transparency, provider duties, and the prohibited-practice ban.

None of that is US law. But three pieces of it are good practice anywhere, and any firm serving EU clients is already in scope. Here is what is worth copying.

1. Tell people when they are talking to AI

The EU rule is blunt: a chatbot or voice assistant has to tell the user it is AI, not a human. Copy it, because the trust math is one-sided. A prospect who discovers halfway through a conversation that the confident, helpful voice was a bot does not feel served. They feel handled. That feeling is expensive for a firm whose entire brand is trust.

The fix costs almost nothing. Open every AI chat or voice interaction with a short, plain line: you are talking with our AI assistant, and a person is one step away. Keep a clean human handoff for anything sensitive. We build that disclosure into every chatbot and voice agent we ship, because it is the cheapest trust you will ever buy, and for clients with any EU footprint it is now required rather than optional.

2. Label your AI-generated content

The EU now requires that AI-generated or altered images, video, and audio be labeled, and that AI content carry machine-readable marks so it can be detected. More than 180 organizations have already signed the Commission's Code of Practice on transparency of AI-generated content. The direction is clear, and the tools to detect unlabeled AI content are only getting better.

For a reputation-driven firm, the lesson is not to hide that you use AI. It is to be visibly honest about it. Label AI-generated visuals. Keep provenance information on the assets you publish. Where it matters, say plainly that a piece was AI-assisted and reviewed by a person. Being the firm that discloses beats being the firm that gets caught, and the gap between those two positions is a one-time process change, not a budget line.

3. Keep a paper trail, and skip the manipulative stuff

Two more pieces of the AI Act point the same way. Providers of the largest AI models now have to document what their models do and summarize the data used to train them. And the prohibited-practice ban outlaws systems that manipulate people, exploit vulnerabilities, or score them unfairly. The documentation duty falls on the frontier labs, not on your firm. The lesson for your firm is the habit underneath it.

Keep your own governance evidence. A short AI register is enough to start: each AI system you run, what it does, what data it touches, who owns it, and when it was last reviewed, plus the compliance posture it lives under, whether that is bar advertising rules, HIPAA, or the SEC Marketing Rule. Then steer clear of the manipulative design the EU is banning: no fake scarcity, no dark patterns, no AI that pressures a vulnerable client. When a client, a board, or a regulator asks how your AI works and whether it is safe, evidence is the answer, and the firm that already has it looks like the adult in the room. Building and maintaining that trail is the heart of our Fractional CAIO work.