From 2 August 2026, transparency under the EU AI Act became a live compliance issue: AI-generated or manipulated text, images, audio and video must be identifiable to users where the Regulation applies to the specific system and role. For a Georgian company, the key question is not only where the business is incorporated. It is also what the company supplies, who uses it and whether the activity connects to the EU market.

What changed in August 2026

The European Commission states that the Article 50 transparency duties apply from 2 August 2026. Regulation (EU) 2026/1744 gives providers of generative AI systems already placed on the market before 2 August 2026 until 2 December 2026 to adapt their practices for marking AI-generated content.

This does not mean every Georgian company automatically falls under every chapter of the AI Act. Scope depends on the system, the company’s role — provider, deployer, importer or distributor — and the product’s connection to the EU market. The practical answer is an inventory of each AI tool, not a general compliance slogan.

A four-step review for Georgian businesses

List every AI system in use: chatbots, advertising generators, HR screening, document analysis and customer profiling.
Identify the role and market: who develops the system, who controls it and whether its output serves EU users.
Obtain vendor documentation, content-marking methods, logs, security controls and incident-notification procedures.
Put data use, confidentiality, model changes, audit rights, indemnity and termination rules into the contract.

If the same risk arose in Georgia, which laws would matter

In Georgia, the Law on Personal Data Protection would become central where the system processes an identifiable person’s data, behavioural profiles, biometric data or employment information. The company should document purpose, legal basis, minimisation, security, notice and data-subject rights.

Contract law would allocate responsibility between the Georgian business and the AI vendor. Copyright issues may also arise from training data, generated material and third-party content. For high-impact decisions, a company should preserve meaningful human oversight and a route to challenge the outcome; an automated answer should not become a substitute for accountable decision-making.

What management should do before 2 December

Start with an AI register and risk matrix. Then test marking in every customer-facing language, review supplier agreements and complete a data-protection impact assessment for projects that could materially affect people. A short, readable public explanation is often more useful than a purely technical policy.

Frequently asked questions

Does the EU AI Act apply to a Georgian company? It may, if the company places an AI system on the EU market, acts in a relevant role or supplies output used in the EU; the answer depends on the facts.
Is the label “AI-generated” always enough? Not necessarily. The specific technical and user-facing marking requirement must be assessed.
Do Georgian laws replace the AI Act? No. Georgian data-protection and contract rules create separate duties, while an EU market connection can add the AI Act regime.

Primary sources: EUR-Lex, Regulation (EU) 2026/1744 — https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32026R1744 ; European Commission, AI-generated content marking and labelling — https://digital-strategy.ec.europa.eu/en/news/commission-publishes-code-practice-marking-and-labelling-ai-generated-content ; Georgian Law on Personal Data Protection — https://www.matsne.gov.ge/en/document/view/5827307

NoticeThis publication is for general information and is not individual legal advice.