Adversa AI Releases 2025 AI Security Incidents Report

Adversa AI has unveiled its 2025 AI Security Incidents Report, highlighting significant vulnerabilities in generative and agentic AI systems, announced in a press release. The report reveals that AI systems, from chatbots to autonomous agents, are increasingly being exploited, with incidents doubling since 2024.

Key findings indicate that 35% of AI security incidents were caused by prompt injections, leading to substantial financial losses. While generative AI was involved in 70% of incidents, agentic AI caused the most severe failures, including crypto thefts and API abuses. The report also notes that breaches often stem from improper validation and infrastructure gaps.

The report includes 17 real-world case studies and provides actionable guidance for improving AI security. Adversa AI, a leader in AI Red Teaming and Agentic AI Security, continues to protect organizations by identifying vulnerabilities before they reach production. For more information, visit Adversa AI.