Kaspersky warns surge in AI‑tool malware attacks on SMBs

Kaspersky's latest security report recorded more than 33,300 cyber‑attacks worldwide between January and April 2026 that targeted small‑ and medium‑size businesses (SMBs) through malicious software disguised as popular AI services such as ChatGPT, Claude and DeepSeek. The number of incidents rose almost five‑fold compared with the same period in 2025.

In Southeast Asia alone, Kaspersky detected over 1,800 attacks in the first four months of 2026, a rise of nearly seven times year‑on‑year. Fake AI tools were the most common lure, accounting for 44 % of the malicious downloads, followed by DeepSeek (33 %) and Claude (11 %). The majority of the disguised files are trojware variants that can steal, modify, delete, lock, or copy user data and may open the way for further intrusion.

The firm urges users to obtain AI applications only from official sources and to use up‑to‑date security solutions to mitigate the growing threat.