The government warned that personal information leaked in last year's successive large-scale hacking incidents could lead to "secondary damage" this year, such as voice phishing and smishing, urging extra caution. As everyday chatbot and generative artificial intelligence (AI) services become targets of attacks, it expects greater risks of malfunctions and information exposure.
The Ministry of Science and ICT and the Korea Internet & Security Agency (KISA) on the 27th released the report "2025 cyber threat trends and 2026 outlook." Twelve domestic and overseas information security corporations, including AhnLab and Genians, took part in the analysis.
Cyber incident reports last year totaled 2,383, up 26.3% from the previous year (1,887). The growth rate was higher in the second half (36.5%) than in the first half (15%), concentrating incidents in the latter half. Ransomware accounted for 11.5% (274 cases) of the total, with a similar share, but the number of cases rose 10.3%.
The report cited the expansion of AI-based attacks, abuse of discontinued or neglected IT services, exploitation of cloud vulnerabilities, and sophisticated scams by combining large-scale leaked data as 2026 threats. It said that if information leaked in last year's SKT, KT, and Coupang incidents is combined, deepfake voice and video phishing could spread to real-time calls and video conferences.
It also raised the possibility of targeting AI services themselves, such as ChatGPT and Gemini, by injecting malicious prompts or tampering with training data to induce unintended behavior and information leakage. As cloud migration accelerates, it urged continuous checks of accounts, access privileges, and security settings. With Windows 10 support ending in October this year, it warned that attacks exploiting management gaps could increase and emphasized following basic rules such as updates, multi-factor authentication, and blocking suspicious links.