The EU AI Act and LinkedIn outreach: why Elvatix is a copilot.
The transparency rules of the EU AI Act have applied since 2 August 2026. What do they mean for AI in your recruitment messages, and why did Elvatix deliberately choose a copilot where you send the message yourself?
The EU AI Act does not ban AI in recruitment messages. Since 2 August 2026, an AI system that interacts directly with people must disclose that it is an AI (Article 50). Elvatix is a copilot: you review every message and send it yourself, from your own LinkedIn account and in your name.
- Elvatix puts no bot in your browser clicking around on your behalf; everything goes out from your own account. Your account safety
- Every candidate shows the reasons for the score, including why someone drops out. AI sourcing
- You review your InMails and connection requests before they go out. FAQ
What the law says, and what it means for you.
The AI Act does not ban AI in your messages
The EU AI Act (Regulation (EU) 2024/1689) does not prohibit using AI to write recruitment messages. It sets rules on transparency and, for certain uses, on risk management and human oversight.
An AI that communicates by itself must say so
Under Article 50(1), an AI system intended to interact directly with people must be designed so those people know they are dealing with an AI. This has applied since 2 August 2026, for example to chatbots and agents that hold conversations on their own. Infringements can be fined up to €15 million or 3% of worldwide annual turnover (Article 99).
Why Elvatix is a copilot
In Elvatix the AI does not hold a conversation with the candidate. Elvatix writes a draft; you review your InMails and connection requests, change what you want and send them from your own LinkedIn account, in your name. The sender is a person. The follow-up after an accepted connection request is something you switch on per list; it goes out from your account on the day you choose, and if the conversation is already running, nothing happens.
Do you have to say that AI helped write it?
Article 50(4) requires disclosure for AI-generated text published to inform the public on matters of public interest, unless a person has reviewed the text and holds editorial responsibility for it. A personal message to one candidate is not a publication for the public, and in Elvatix you review every message you send. In our reading, as a recruiter you therefore do not need to add that AI helped write it.
AI in recruitment and selection: what changes on 2 December 2027
The AI Act lists AI for recruitment and selection, such as evaluating candidates, as a high-risk use (Annex III, point 4(a)). After the Digital Omnibus decision (Regulation (EU) 2026/1744), the requirements apply from 2 December 2027, including risk management, documentation, transparency and human oversight. Elvatix is built around human oversight: every candidate shows the reasons for the score, including why someone drops out, and you decide who to approach.
What this means for fully automated tools
A tool that sends messages and holds conversations on its own interacts directly with the recipient as an AI. The disclosure duty of Article 50(1) then applies. The candidate then knows they are talking to an AI. With a copilot it stays your message: personal, in your name, and with the response rate that comes with it, such as 43% at Manpower and consistently above 50% for Kristel Frey.
Sources
- Regulation (EU) 2024/1689 (AI Act), Article 50: transparency
- AI Act, Annex III: high-risk AI applications
- AI Act, Article 99: penalties
- Regulation (EU) 2026/1744 (Digital Omnibus on AI), Official Journal of the EU
- Council of the EU, press release 29 June 2026
Read on 17 October 2026. This page is an explanation, not legal advice; if in doubt, consult a lawyer.
Questions about the AI Act.
Yes. The AI Act does not prohibit using AI to write recruitment messages. It sets transparency rules: an AI system that interacts directly with people must make clear that it is an AI (Article 50(1)).
In our reading, no, as long as you review the message and send it yourself. The disclosure duty for AI text in Article 50(4) concerns publications for the public on matters of public interest, and it falls away with human review and editorial responsibility. A personal message to one candidate is not such a publication.
The transparency rules of Article 50 have applied since 2 August 2026. After the Digital Omnibus decision, the requirements for high-risk AI in recruitment and selection apply from 2 December 2027.
A tool that sends messages and holds conversations on its own interacts with the recipient as an AI and must therefore disclose it. With a copilot the AI writes a draft and a person sends the message; the sender is then a person.
For breaching the transparency rules of Article 50, up to €15 million or 3% of worldwide annual turnover (Article 99). For SMEs and start-ups, the lower of the two applies.
AI that writes, a person who sends.
See on your own vacancy how Elvatix writes personal messages that you review and send yourself.
You're not on your own: the AI assistant in Elvatix is with you at every step. How the assistant works