First step towards market regulation artificial intelligence made by Europe European Parliament Last week, a corresponding bill was approved, which could become the first law on artificial intelligence in the West.

The proposed rules are based on the risk of AI applications, i.e. the obligations of the system arise in accordance with the risks they cause. Specific specifications are also set for so-called basic models such as models ChatGPT which are of great concern as they may even result in the reduction of skilled jobs.

Specifically, the bill breaks down AI into four levels of risk: unacceptable, high, limited, and low or no risk. Applications with unacceptable risk include those that parse or use sensitive information or use deception techniques to influence human behavior. And, of course, they are proposed to be completely banned in the European bloc.

At the same time, however, after repeated calls from officials, there is agreement on a foundation model, a wide network of applications, including ChatGPT. As part of these rules, developers must perform security checks, data management measures, and risk mitigation measures before providing access to the public. You should also make sure that the data used to update them does not violate intellectual property rules.

In recent months, major companies technologies Various artificial intelligence applications, mainly chatbots, have become available to the general public, as a result of which the market is growing at a rapid pace.

Just last week, for example, Google announced a series of AI updates, including the PaLM 2 language model, which the company says is better than other related apps from competing companies.

In general, new chatbots such as ChatGPT have impressed experts and analysts with their ability to provide answers in an almost human style. On the contrary, artificial intelligence technologies in general have been developed over the years and are already being used in many systems. Ultimately, the goal of the EU proposals is to introduce certain rules for the further development of AI companies, as well as for organizations that use it.