“The prime minister tasked me with reviewing what competencies we have and what we can do in the field of artificial intelligence, as a lot is happening in Hungary in this area,” said Laszlo Palkovics, government commissioner for AI. He emphasized that coordinating AI-related activities is important, and the government plays a significant role in this, for example, in regulating that a single shared tool is sufficient for state applications.
We have technological and awareness-raising tasks; we have to tell people that AI, which is now much more than just a technology, is not dangerous but comes with many advantages. Our job is to make it as efficient as possible, which we discuss with universities, research institutes, state institutions and companies,
the commissioner said.
Easier Access to Information
In his view, everyone has some connection to AI, as most people already use ChatGPT-type functions. It is not surprising that large multinational companies use AI in their manufacturing and HR processes, just as it is not surprising that banks have long operated with AI-based solutions when assessing their customers’ loan applications. The Ministry of Justice has been working for years on how to "teach" the Hungarian legal database, as well as court and municipal decisions to AI, as this enables easier access to information. “There is no area where we do not use AI, we just do so with differing levels of intensity and depth,” Palkovics emphasized.
He added that
another major advantage of the rapidly developing AI is that it can detect patterns in data that previous statistical methods could not.
Using past information, they can create control algorithms and find new solutions. Preparations are underway to develop technology that will later be able to determine whether AI was used to create certain content, he explained.
Those Using AI Illegally Will Suffer the Consequences
The commissioner cited the unlawful use of AI to spread fake news during election campaigns as a relevant risk, for example using deepfake technology to commit forgery, falsify videos, or put lies into the mouths of certain public figures.
I would caution everyone against using this in election battles, as it could unlawfully enable AI to influence election results. The EU AI Act requires that if content is created using AI, this must be clearly indicated, as otherwise it misleads others, causes them harm, or infringes on their personal rights,
he warned.
Laszlo Palkovics highlighted that one of the most challenging issues regarding AI is copyright. This is not about someone copying an article or a book and selling it as their own but about someone reading a book they wish to understand better and using AI for this purpose. “The question is whether I can buy this book and, if I have the capability, find correlations that I can then use in my own work. There is currently a major debate on this: if there are expired patents from which I create a new patent, who does that belong to?” he noted.
Promising Applications in the Automotive Industry
The commissioner pointed out that AI is widely used in car manufacturing in multiple ways: in product development, analyzing manufacturing processes, quality management using image recognition algorithms, and in testing processes. Moreover, this is only the device-side application; AI can also be integrated into the product itself, with good examples being various parking functions and driver behavior monitoring. “The big challenge here is how to make these safe.”
We developed and continue to improve the test track in Zalaegerszeg so we can test scenarios that may occur with self-driving cars.
If an autonomous vehicle does not react appropriately in a traffic situation, it can be retrained, but many tests are still needed to ensure that these vehicles can operate accident-free in everyday use,” said Laszlo Palkovics.
