cyber.hendryadrian.com
Security Report · darktrace.com

Darktrace Survey: 87% Say AI Is Increasing Malware Sophistication

· Original article ↗

Summary

Darktrace’s 2026 survey finds widespread concern about AI-driven threats: 87% of respondents say AI increases malware sophistication, while 46% feel inadequately prepared to defend their organizations.

Key points

  • Nearly 62% of surveyed organizations reported a social engineering attack involving a deepfake or AI-generated audio or video used to target biometric authentication; 32% reported an AI prompt-injection attack.
  • Respondents said attackers use AI across the attack chain, including reconnaissance, tailored phishing, lateral movement, and data exfiltration.
  • 73% said AI-powered cyber threats significantly affect their organization, and 92% said the threats are forcing them to upgrade defenses.
  • 87% said AI significantly increases malware sophistication and success rates; the same share said it increases security operations workload.
  • Hyper-personalized phishing was a major concern for 50% of security leaders; 45% cited automated vulnerability scanning and exploit chaining, and 40% cited deepfake voice fraud and adaptive malware.
  • 46% of security leaders said their organizations were inadequately prepared, similar to 45% the previous year; confidence varied by country.
  • The report identifies insufficient knowledge or use of AI-driven countermeasures as a key defensive gap; Darktrace also promotes its own AI security approach.

Article Details

Publisher
Darktrace
Report Period
2026
Scope
Survey findings on AI-driven cyber threats, security professionals' concerns, and organizational preparedness; also includes observations of malicious emails in 2025.
Key Statistics
  • Nearly 62% of organizations experienced a social engineering attack involving a deepfake or AI-generated audio or video used in an attempted biometric-authentication deception; 32% reported an AI prompt injection attack.
  • 73% said AI-powered cyber threats were having a significant impact on their organization, and 92% agreed these threats were forcing them to upgrade defenses.
  • 87% said AI was significantly increasing the sophistication and success rate of malware; 87% also said AI was significantly increasing their security operations team's workload.
  • 46% of security leaders felt inadequately prepared to protect their organizations, compared with 45% in the prior year and 60% in 2024.
  • 33% of the malicious emails Darktrace observed in 2025 contained over 1,000 characters, which the article says indicates probable LLM usage.

Vendors

Anthropicpersonalize phishing, accelerate lateral movement, and automate data exfiltration. Evidence from Anthropic demonstrates that threat actors have harnessed AI to orchestrate an entire cyber espionage campaign fromDarktraceThe findings in this blog are taken from Darktrace’s annual State of AI Cybersecurity Report 2026.

Countries

BrazilSome of these differences are accentuated across different cultures. Respondents in Japan are far less confident (77% say they are not adequately prepared) than respondents in Brazil (where only 21% don’t feel prepared).JapanSome of these differences are accentuated across different cultures. Respondents in Japan are far less confident (77% say they are not adequately prepared) than respondents in Brazil (where only 21% don’t feel prepared).

Related Articles