All practice areas

ARTIFICIAL INTELLIGENCE LAW LAWYER IN PLOVDIV

Artificial intelligence law (AI Act)

Lawyer Miroslav Mirchev is a lawyer in Plovdiv who specialises in artificial intelligence law. He advises companies that develop, deploy or use AI systems on the application of Regulation (EU) 2024/1689 (the Artificial Intelligence Act): determining the organisation’s role, risk classification, transparency obligations, internal rules for the use of generative AI and contracts with AI solution providers.

The regulation of artificial intelligence in the EU is being introduced in stages and is already mandatory in a significant part. The prohibited practices have applied since February 2025, the obligations for general-purpose models since August 2025, and from 2 August 2026 the general transparency requirements apply: users must know that they are communicating with artificial intelligence. The obligations for high-risk systems have been postponed and will apply from 2 December 2027 for stand-alone systems and from 2 August 2028 for systems embedded in regulated products.

These services are most often needed by start-ups and companies that integrate artificial intelligence into their product or their work: through a chatbot, an AI assistant, their own or an external model, or as a component in software they offer to clients. For them, regulation is not only a future risk. It already determines how the contract with the developer is written, what is promised to clients and what internal rules are needed.

Two questions arise most often. The first is what obligations the AI Act places on a specific organisation: they depend on whether it is a provider or a deployer of the system, and on the risk class. The second concerns contracts for the development of AI systems: who is liable for errors and damage, who owns the data, the model and the outputs, and what guarantee is given of compliance with the regulation.

Правно регулиране на изкуствения интелект и AI Act

Services in the field of artificial intelligence

Artificial intelligence law (AI Act)

Legal advice on artificial intelligence

Compliance with the AI Act, liability and rights in generated content.

Artificial intelligence law (AI Act)

Determining your role under the Artificial Intelligence Act

Provider, deployer, importer or distributor: your role determines the whole scope of your obligations under the AI Act.

Artificial intelligence law (AI Act)

Risk classification of an AI system

Prohibited practice, high risk, limited risk or minimal risk: determining the applicable regime.

When you need an artificial intelligence law lawyer

  • when you build a chatbot or an AI assistant into a website, an app or customer service
  • when you use generative AI to create published content
  • when you develop or sell software with an AI component
  • when you conclude a contract with an AI solution provider
  • when your employees already use AI tools but you have no internal rules

Frequently Asked Questions

Does the AI Act apply to my company if I only use ChatGPT?
Yes, but in the lightest regime. An organisation that uses an AI system in a professional activity is a “deployer” within the meaning of the Regulation. The main obligations in this case relate to transparency and to ensuring a sufficient level of AI literacy among employees.
Do I have to disclose that I use AI to create content?
If you publish text, images or video created with generative AI, the general transparency obligation applies to you from 2 August 2026: users must know that they are communicating with artificial intelligence, and artificially generated or manipulated content must be labelled.
What does “high-risk AI system” mean?
The Regulation classifies as high-risk the systems used in areas that directly affect people’s rights and safety, including recruitment and personnel management, access to education, creditworthiness assessment, critical infrastructure, healthcare and law enforcement. AI components embedded in products that are already subject to safety regulation are also high-risk. The classification determines whether the most demanding requirements apply: risk management, technical documentation, human oversight and conformity assessment.
Can I enter client data into an AI tool?
It depends on the legal basis for processing, the terms of the particular provider and whether the data leaves the EU. Entering personal data or trade secrets into a public AI service without prior analysis is one of the most common sources of risk in practice. The solution is usually a combination of internal rules, choosing a suitable service tier and a contract with the provider.
Who owns content created by artificial intelligence?
Bulgarian copyright law recognises as author the natural person who created the work. Content generated entirely automatically is hard to classify as a protected work, which means that you too may not have exclusive rights in it.
What obligations does the AI Act place on a company that integrates artificial intelligence into its product?
It depends on the role: provider (who develops the system or places it on the market under its own name) or deployer (who uses it). Providers have broader obligations. First of all, the role must be determined and the risk of the system classified.
What should a contract for the development of an AI system contain?
A definition of the system and the expected results, rights in the code, the model and the training data, liability for errors and damage, guarantees of compliance with the AI Act and personal data protection, and, for future changes, maintenance.