Czech Ombudsman Stanislav Krecek told CTK yesterday that he intends to investigate the effect of artificial intelligence (AI) on human rights, adding that it was important that people, not AI, should answer to people.
The ombudsman’s office itself also uses artificial intelligence.
“I think that artificial intelligence and human rights will be one of the topics that I and my successors will deal with, also taking into account that we are now a national human rights institution,” said Krecek. “The authorities, myself and my team deal with the real problems of real people. That is why it is important that people, and not artificial intelligence, answer to people. The answers must reflect what people are concerned about and must include responses to specific situations and objections, not what the AI has in its database.”
According to Vaclav Rendl of the Ombudsman’s Office, the Ombudsman has not knowingly dealt with any complaints in its agenda related to artificial intelligence. However, he said, this cannot be ruled out, because the use of an AI system may not be obvious at first sight.
However, Rendl also expects the ombudsman to start looking into the issue. “Artificial intelligence will enter all areas of life, and therefore also the ombudsman’s agendas. As the National Human Rights Institution (NHRI), we will be looking a lot at areas where the development and use of AI may interfere with fundamental rights,” Rendl said.
Artificial intelligence is also used by the Ombudsman’s staff. According to Rendl, there are strict rules in place for working with it. “As a matter of principle, we do not put any personal/protected data into AI tools. Staff have also been trained on how to use current AI tools. We have internal citation rules when using AI tools,” Rendl said.
So far, he said, staff in the Ombudsman’s Office use AI more for research work. For example, DeepL helps with translations of technical texts from less common languages, NotebookLM with analysis of technical texts, and ChatGPT to create brief annotations of case law or other summaries. They are also looking for other ways to expand the use of AI tools and make their work more efficient, he said. These include a secure tool for machine anonymisation of texts.
Artificial intelligence has proliferated in recent years thanks to advances in computing and machine learning algorithms. It is increasingly used by companies and government institutions, especially for automation, data analysis and decision support. Language models have also gained ground.
