CrowdStrike Report: AI Accelerates Attacks and Expands the Security Threat Surface

· Original article ↗

Summary

CrowdStrike’s annual report says AI-enabled malicious activity rose 89% over the past year, with AI used against vulnerabilities, software supply chains and AI systems. It reports that 88% of vulnerabilities were weaponized through AI within 48 hours.

Key points

  • CrowdStrike reports AI-enabled malicious activity rose 89% during the year covered by its report.
  • The company’s threat-hunting team and systems triaged an average of 14 million detection leads daily, resulting in about 36,000 customer alerts over the year ending in June.
  • The report says attackers use frontier AI models to find vulnerabilities and generate scripts, payloads and commands.
  • CrowdStrike says 88% of vulnerabilities were weaponized through AI within 48 hours; Adam Meyers said this challenges organizations’ 30-day patch cycles.
  • Meyers said the TeamPCP cluster compromised more than 300 software dependencies in one day.
  • The report warns that AI tools, agentic systems and integrations are expanding the attack surface, while many organizations have not put adequate safeguards around their AI tools.

Article Details

Publisher
CrowdStrike
Report Period
One-year period ending in June; the year is not specified in the supplied text.
Scope
CrowdStrike annual threat hunting report covering AI-enabled activity, vulnerability weaponization, open-source software supply-chain attacks, and threats to AI systems.
Key Statistics
  • CrowdStrike triaged an average of 14 million detection leads daily during the one-year period ending in June.
  • The leads resulted in about 36,000 customer alerts during that period.
  • CrowdStrike reported that AI-enabled malicious activity rose 89% over the past year.
  • CrowdStrike reported that 88% of vulnerabilities were weaponized through AI within 48 hours.
  • CrowdStrike's Adam Meyers said TeamPCP compromised more than 300 software dependencies in one day.
Recommendations
  • Secure AI tools and put appropriate guardrails around their use.
  • Organizations should account for a 24-to-48-hour patch cycle, which Meyers said is the new baseline.

People

Threat Actors

Related Articles