AI Security Point

Newsletters / Monthly

AI Security Monthly — May 2026

May 1, 2026 – May 31, 2026 · generated Sep 10, 2026

Total Incidents30
Previous Period25
Period-over-Period Growth+20.0%
Annual Run Rate392
Annual Run Rate Growth+2.1%
Quarterly Run Rate80
Quarterly Run Rate Growth-27.9%

Breakdowns

By Type
  • ML18
  • LLM7
  • Agent3
  • Identity2

All machine-classified (keyword-v1), not yet human-reviewed.

By Severity
  • Minor4
  • Severe1
  • Substantial1

6 of 30 incidents in this period (20%) have been classified on this dimension; all machine-classified (mitrisk-navigator), not yet human-reviewed.

By OWASP Top 10
  • ASI01:2026 — Agent Goal Hijack2
  • LLM01:2025 — Prompt Injection2
  • LLM05:2025 — Improper Output Handling2
  • LLM09:2025 — Misinformation2
  • ASI02:2026 — Tool Misuse and Exploitation1
  • ASI03:2026 — Identity and Privilege Abuse1
  • ASI07:2026 — Insecure Inter-Agent Communication1
  • LLM02:2025 — Sensitive Information Disclosure1
  • LLM06:2025 — Excessive Agency1

7 of 30 incidents in this period (23%) have been classified on this dimension; all machine-classified (framework-agent-v3), not yet human-reviewed.

By MIT AI Risk Domain
  • Malicious Actors & Misuse13
  • 4.3. Fraud, scams, and targeted manipulation5
  • AI System Safety, Failures, & Limitations3
  • 2.2. AI system security vulnerabilities and attacks2
  • 3.1. False or misleading information2
  • 7.3. Lack of capability or robustness1
  • Discrimination & Toxicity1
  • Misinformation1
  • Privacy & Security1
  • Socioeconomic & Environmental1

All machine-classified (framework-agent-v3, mitrisk-navigator), not yet human-reviewed.

Incidents this period (30)

Classification fields only. Follow each ID to the AI Incident Database for the full record and source reports.

IncidentDateDomainMIT AI Risk
AIID #1508May 31, 2026ML4.3. Fraud, scams, and targeted manipulation
AIID #1510May 31, 2026LLMPrivacy & Security
AIID #1512May 31, 2026AgentMisinformation
AIID #1507May 30, 2026LLM2.2. AI system security vulnerabilities and attacks
AIID #1531May 30, 2026MLMalicious Actors & Misuse
AIID #1593May 28, 2026MLSocioeconomic & Environmental
AIID #1505May 24, 2026ML4.3. Fraud, scams, and targeted manipulation
AIID #1582May 24, 2026MLMalicious Actors & Misuse
AIID #1666May 22, 2026MLAI System Safety, Failures, & Limitations
AIID #1584May 20, 2026MLDiscrimination & Toxicity
AIID #1501May 19, 2026LLM4.3. Fraud, scams, and targeted manipulation
AIID #1590May 19, 2026MLMalicious Actors & Misuse
AIID #1530May 16, 2026MLMalicious Actors & Misuse
AIID #1503May 15, 2026ML7.3. Lack of capability or robustness
AIID #1524May 15, 2026MLMalicious Actors & Misuse
AIID #1523May 14, 2026MLMalicious Actors & Misuse
AIID #1497May 13, 2026Agent2.2. AI system security vulnerabilities and attacks
AIID #1522May 13, 2026MLMalicious Actors & Misuse
AIID #1504May 12, 2026LLM3.1. False or misleading information
AIID #1520May 11, 2026MLMalicious Actors & Misuse
AIID #1521May 11, 2026MLMalicious Actors & Misuse
AIID #1599May 9, 2026AgentAI System Safety, Failures, & Limitations
AIID #1499May 6, 2026LLM3.1. False or misleading information
AIID #1662May 6, 2026MLAI System Safety, Failures, & Limitations
AIID #1485May 5, 2026LLM4.3. Fraud, scams, and targeted manipulation
AIID #1519May 5, 2026MLMalicious Actors & Misuse
AIID #1480May 4, 2026ML4.3. Fraud, scams, and targeted manipulation
AIID #1556May 4, 2026LLMMalicious Actors & Misuse
AIID #1534May 1, 2026IdentityMalicious Actors & Misuse
AIID #1579May 1, 2026IdentityMalicious Actors & Misuse