{"id":381682,"date":"2026-10-21T08:24:00","date_gmt":"2026-10-21T15:24:00","guid":{"rendered":"https:\/\/cms-articles.softonic.io\/en\/?p=381682"},"modified":"2026-10-21T08:24:08","modified_gmt":"2026-10-21T15:24:08","slug":"crowdstrike-exposes-four-threat-groups-language-models-speed-up-break-ins","status":"publish","type":"post","link":"https:\/\/cms-articles.softonic.io\/en\/crowdstrike-exposes-four-threat-groups-language-models-speed-up-break-ins\/","title":{"rendered":"CrowdStrike exposes four threat groups: language models speed up break-ins"},"content":{"rendered":"

CrowdStrike has published new research on four threat groups<\/strong> it says hadn’t been reported before. Between May and August, those groups used language models to speed up familiar attacks, and CrowdStrike traced the operators to Chinese, Spanish\/English, Indonesian, and Russian environments.<\/p>\n

\r\n
\r\n
\r\n
\r\n \"Avast\r\n <\/div>\r\n
\r\n Avast Free Antivirus<\/span>\r\n Download<\/a>\r\n <\/div>\r\n
\r\n <\/path><\/path><\/text><\/svg>\r\n <\/div>\r\n <\/div>\r\n
\r\n <\/span>\r\n <\/div>\r\n
\r\n \r\n <\/div>\r\n <\/a>\r\n <\/div>\r\n<\/div>\n\n\n\n

The trail started with exposed public servers. From there, CrowdStrike says the Chinese and Spanish\/English<\/strong> groups built multi-step harnesses that could run for long stretches with very little human input. The Indonesian and Russian groups weren’t as advanced, but they still went after multiple targets. All four managed to get around safeguards and mainly relied on Anthropic Claude and open-weight DeepSeek<\/a> models, often older versions picked for capability, lower cost, and fewer restrictions.<\/p>\n\n

CrowdStrike says the operators leaned on jailbreaks, fake penetration-testing narratives, and planted approval files. The bigger issue for defenders is that they didn’t need any new exploits<\/strong>. According to CrowdStrike, the models let autonomous campaigns move from reconnaissance to lateral movement in minutes, going after poor hygiene, exposed services, and misconfigurations that still hadn’t been fixed. That lines up with other data CrowdStrike cites: 80% of phishing is machine-generated, these tools showed up in 84% of ransomware attacks in 2026, and nearly nine in ten organizations were hit last year.<\/p>\n\n

If you manage security, this CrowdStrike report is worth reading. Basic hygiene still matters, but now it has to be backed by faster automated defenses<\/a><\/strong> and quicker incident sharing from model developers. You can find the report on CrowdStrike’s website.<\/p>\n

\r\n
\r\n
\r\n
\r\n \"Malwarebytes\r\n <\/div>\r\n
\r\n Malwarebytes Anti-Malware<\/span>\r\n Download<\/a>\r\n <\/div>\r\n
\r\n <\/path><\/path><\/text><\/svg>\r\n <\/div>\r\n <\/div>\r\n
\r\n <\/span>\r\n <\/div>\r\n
\r\n \r\n <\/div>\r\n <\/a>\r\n <\/div>\r\n<\/div>\n\n","protected":false},"excerpt":{"rendered":"

CrowdStrike has published new research on four threat groups it says hadn’t been reported before. Between May and August, those groups used language models to speed up familiar attacks, and CrowdStrike traced the operators to Chinese, Spanish\/English, Indonesian, and Russian environments. The trail started with exposed public servers. From there, CrowdStrike says the Chinese and … Continue reading “CrowdStrike exposes four threat groups: language models speed up break-ins”<\/span><\/a><\/p>\n","protected":false},"author":9349,"featured_media":381681,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":"","wpcf-pageviews":0},"categories":[1015],"tags":[],"usertag":[],"vertical":[],"content-category":[6771],"class_list":["post-381682","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-news","content-category-ai"],"aioseo_notices":[],"aioseo_head":"\n\t\t\n\t\n\t\n\t\n\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t\n\t\t