Rapid7: High-Severity Flaws Exploited in Three Days as Ransomware Leads Q3 Threats

· Original article ↗

Summary

Rapid7 reports that high-severity flaws were exploited within three days on average in Q3 2026, while ransomware remained dominant. AI appeared in 8% of tracked incidents, with humans still directing targets and objectives.

Key points

  • High-severity vulnerabilities were exploited an average of three days after disclosure; 71% were exploited within seven days.
  • Rapid7 classified 79% of the high-severity vulnerabilities as remote flaws requiring neither credentials nor local user involvement.
  • Ransomware activity most often targeted business services, manufacturing, and healthcare. The US accounted for 60% of attacks tracked.
  • The Gentlemen and Qilin led reported ransomware incidents, with 342 and 339 claimed incidents, respectively.
  • AI was cited as an operational component in 8% of tracked incidents. Humans chose targets and objectives in every documented case, though some tactical tasks were automated.
  • Rapid7 described a previously undocumented DPRK-linked Linux toolkit used against South Korean organizations since early 2025, including a backdoor, SSH keylogger, and curl-based RAT.
  • Forty-one of 63 underground posts advertised RDP or network access, which Rapid7 says can provide defenders with early warning of potential attacks.

Article Details

Publisher
Rapid7
Report Period
Q3 2026
Scope
Global threat activity, including vulnerability exploitation, ransomware, AI use, nation-state and cybercrime activity, and dark-web listings.
Sample Size
15 nation-state and cybercrime groups tracked; the total number of incidents is not disclosed.
Key Statistics
  • High-severity vulnerabilities were exploited an average of three days after disclosure; 71% were exploited within seven days.
  • Rapid7 classified 79% of the vulnerabilities discussed as “holy grail” flaws: remote and requiring neither credentials nor local user involvement.
  • The US accounted for 60% of attacks, followed by Germany at 7%, Italy at 5%, and Canada at 5%.
  • AI was cited as an operational component in 8% of tracked incidents; the report says humans chose what to attack and why in every documented case.
  • 41 of 63 underground posts advertised RDP or network access; 14 involved the sale, sharing, or discussion of exploits and zero-days.
Recommendations
  • Monitor AI API usage patterns from unexpected hosts and processes as a source of telemetry, and look for unusual or unauthorized use of AI tools.

MITRE ATT&CK

People

Threat Actors

Malware

Vendors

Products

Countries

Industries

Related Articles