Apr 01, 2021
The March issue of Svet osiguranja, an insurance magazine for Serbia, Montenegro and Macedonia, discussed AI system failures and the necessity of new insurance products that will cover this type of risk. Ivan Divic, Country manager of Adacta Serbia, spoke to the magazine in the article entitled “What if the Chatbot Breaks Down”:
“Since AI is not conscious and cannot be liable, the burden falls on users - operators and/or manufacturers. The combination of the existing regulations in the field of product safety and liability insurance does not adequately cover all the complexity of cases arising from new technologies. Therefore, it is important to approach this issue systematically to ensure the technology is developed safely,” said Divic.
“There are still no insurance policies that fully cover the risks brought on by AI. There is a cyber insurance policy, which can cover theft of machine learning models, cyber extortion, data security breaches and privacy, and liability for professional technology services. What is still not included is the coverage for material and non-material damages caused by system failure. Insurers that monitor new forms of risk should certainly provide this type of insurance in their offer, along with risk management services.”
The article was published in Svet osiguranja (The World of Insurance), a professional and market-oriented, independent magazine for law, economics and practice in insurance and reinsurance. It has a regional character and is distributed to insurance companies in Bosnia-Herzegovina, Montenegro, Croatia, Macedonia, Slovenia and Serbia.