Regulation (EU) 2024/1689 – the AI Act – is frequently discussed in Georgia as European news. It is not only European news, for two reasons that matter directly to legal practice here.
Justice-sector tools are treated as high-risk
The Act classifies AI systems intended to assist judicial authorities in researching and interpreting facts and law, or in applying law to facts, as high-risk. That classification carries substantive obligations rather than disclosure alone: risk management, data governance, technical documentation, logging, transparency, accuracy and robustness requirements, and meaningful human oversight.
The human-in-the-loop requirement deserves particular attention, because it is easy to satisfy on paper and hard to satisfy in fact. Oversight that consists of a professional approving output they have no practical capacity to evaluate is not oversight. Making it real requires that the reviewing lawyer understand the system well enough to disagree with it – which is a training question before it is a compliance question.
Extraterritorial reach
The second reason is scope. The Act applies to providers placing systems on the Union market and, importantly, where the output of a system is used within the Union – irrespective of where the provider is established. A Georgian legal-technology provider serving clients or matters in the EU is therefore inside the regime, not adjacent to it.
For a developing legal-services market this cuts both ways. It imposes a compliance burden on providers who are not resourced for it. It also supplies a ready-made standard: a jurisdiction without its own detailed AI framework can harmonise proactively rather than improvise, and treat the Act as the floor rather than a foreign imposition.
What follows practically
Three things follow, and all three are institutional rather than technical. Academic programmes need to teach the regulatory framework alongside the technology, because a lawyer cannot advise on a system whose mechanism is opaque to them. Professional ethical codes need updating to address technology adoption explicitly. And regulatory harmonisation is better done in advance of a dispute than in response to one.