The South Korean government vowed on Monday to share phishing scam call data with private companies developing artificial intelligence-based software and services to counter voice phishing scams, which have consistently been on the rise over the last few years.

Voice phishing scams, or fraudulent phone calls, are estimated to have caused up to 196.5 billion won ($143 million) in damages in 2023, according to Financial Supervisory Service data from March 2024 -- a 35.4 percent increase from numbers for 2022.

The Personal Information Protection Commission, the Ministry of Science and ICT, the Financial Services Commission, the FSS, the National Forensic Service and the Korea Internet and Security Agency signed a memorandum of understanding on Monday promising to share phishing scam call data, which was initially only used for police investigations, with private companies and mobile carriers developing AI-based technology to counter such crimes.

Under the newly established memorandum, the FSS will continuously provide voice call data it collected through reports made from phishing scam victims to the NFS.

After undergoing de-identification processes for the data, the NFS will then establish a data-sharing system for private companies to use when developing AI-based technologies.

The Science and ICT Ministry and the PIPC also agreed Monday to plan and promote government-led technology development projects focused on developing voice analysis models to curb voice phishing scams.

The new agreement will also make it easier for carriers and private companies looking to develop AI-based services and software to prevent voice phishing scams, as it previously had to get in contact with each governmental organization separately to make its project available for release.

According to the PIPC, it has received several requests from mobile carriers and companies to approve its request to develop an AI service to prevent voice phishing scams. Later this month, Korean mobile carrier SK Telecom plans to release an AI service to determine in real time whether a suspicious call is being made based on the context of the call, and to alert the user or a family member.