BTA: Between AI and Regulation: Will Autonomous Vehicles Soon Appear On Bulgarian Streets?

In early March, Associated Press reported that London’s streets are expected to welcome robotaxis soon, while in mid-February dpa said that German carmaker Volkswagen is preparing to launch self-driving minivans in a third European city and its first outside Germany. At the end of 2025, Toyota opened an experimental “Woven City” in Japan featuring autonomous cars and robots.

But where does Bulgaria stand in this process and will autonomous vehicles soon appear on Bulgarian streets?

In response to a BTA inquiry about the development of intelligent transport systems (ITS) in Bulgaria, the Ministry of Transport and Communications said the country is at a key stage in deploying these technologies. The requirements of Directive 2010/40/EU, which establishes a framework for the development of ITS in Europe, have been implemented, along with its amendment – Directive (EU) 2023/2661, which prepares Member States for connected and automated mobility, including autonomous vehicles.

According to the directive, cooperative intelligent transport systems (C-ITS) need to be developed, enabling vehicles to “communicate” with each other and with road infrastructure. This will allow drivers and autonomous vehicles to receive information and warnings about road hazards, handle more complex traffic situations, and facilitate safe autonomous driving.

The Ministry noted that amendments adopted in 2025 to the Ordinance on the conditions and procedure for the deployment of intelligent transport systems, along with the Public Transport Act, have created the necessary conditions to accelerate the introduction of ITS in the country. Additional primary and secondary legislation is planned to fully regulate cooperative and automated mobility, including autonomous driving.

Timeline for the introduction of autonomous vehicles in Bulgaria

The deployment will take place once economic conditions make the technology viable, Assen Nachkov, a PhD student at the Institute for Computer Science, Artificial Intelligence and Technology (INSAIT) at St. Kliment Ohridski University of Sofia, told BTA. He said it is difficult to give a precise forecast, as the adoption of such technology will have a significant impact on the economic structure of the transport industry.

Earlier in 2026, INSAIT presented DiffSim Trinity (diffsimtrinity.insait.ai), a series of three scientific developments that could change the way autonomous vehicles are trained to understand the world and make decisions.

Nachkov outlined several steps to accelerate the process: improving road conditions and infrastructure quality; allowing and designating experimental zones for real-world testing; adopting clear laws on liability (who is responsible in case of accidents), insurance, and safety standards. A careful balance must be struck between the practical benefits of mass deployment and potential restrictions due to rare individual incidents, he added.

On the demand side, factors include public trust in the technology, which can be increased through information campaigns and effective communication. Another important factor is the overall economy. Higher incomes would boost demand, Nachkov said.

According to him, the initial phase of deployment will most likely be driven by private companies that see economic value in entering the market. Among the main challenges for such companies are the significant capital costs for new vehicles, the need to create high-precision maps, adapt and train models to local road conditions, and build a sustainable business model, Nachkov said.

He added that a more substantial challenge is the legislative framework. There is a risk that various groups may oppose the introduction of autonomous vehicles, including through arguments about unfair competition – a scenario already seen with services such as Uber. Such regulatory reactions could significantly delay or even hinder the deployment of the technology, Nachkov noted.

INSAIT and autonomous vehicles

Nachkov said that INSAIT focuses on improving artificial intelligence for vehicle perception and planning capabilities. Perception enables the vehicle to recognize and interpret its surrounding environment – to identify objects and determine their location.

A significant part of the work is dedicated to so-called “differentiable simulation” – a new approach to modelling the behaviour of autonomous vehicles in various situations, he explained. “We have developed some of the largest-scale experiments for training autonomous vehicles using this method, as well as the first artificial intelligence algorithms that use it to improve planning both during training and in real-world deployment,” Nachkov added.

Autonomous or self-driving vehicles are classified as automated vehicles of levels 3, 4 or 5. Level 3 involves conditional automation, Level 4 high automation, and Level 5 full automation.