Data Destruction and Memory Poisoning: Spring AI Vulnerabilities Threaten LLM Integrity

Spring AI released a security advisory detailing three high-severity vulnerabilities—Milvus VectorStore deletion filter-expression injection (data destruction), prompt-memory poisoning in PromptChatMemoryAdvisor (model manipulation), and a cross-user data leakage via a default conversation ID—that affect Spring AI 1.0.x and 1.1.x; maintainers published fixes and advise immediate upgrade to 1.0.7 or 1.1.6.