Explore 3 source-supported AI security cases for 13.34 · Session and State Management Failures in Databricks DASF 3.0, with evidence, scope, and defensive analysis.
Risk category
A client inadequately manages sessions, authentication state, or application state. Attackers can hijack sessions, manipulate state, bypass authentication, or exploit sessions that were not properly terminated. The result can be unauthorized access or privileges being applied to the wrong interaction.
LiteLLM is a gateway that routes applications' requests to language models. Its semantic cache can reuse an earlier answer when two questions have similar meanings. NVISO showed that in the tested configuration, /v1/responses reused cached entries across different API keys. This exposed simulated support data and allowed a fabricated answer or tool-call response to reach a different user. The researchers also demonstrated downstream execution in a mock banking tool and the pi coding agent. CVE-2026-89032 was fixed in v1.101.0-rc.1.
Why it relates: The cache reused a caller’s application and tool state in another caller’s interaction. The observed overlap is the shared runtime cache or inherited gateway exposure in the reported configuration. The test did not change model weights, compromise package distribution or establish takeover of a live MCP account.
All downstream consequences are from separate tests against the vulnerable cache behavior.
pgEdge disclosed independent SQL-injection, read-only-enforcement and bundled-client approval defects affecting versions through 1.0.0 and repaired in 1.1.0. A custom pl-do tool could turn argument data into SQL; query_database could defeat an application-level read-only setting; two client defects skipped write confirmation. Database permissions still bound the resulting actions. The source establishes controlled validation, not a malicious customer breach.
Why it relates: A changed pooled connection can retain writable state and pass that state to later calls under another token. The demonstrated state failure is server-side pooling, while this DASF entry is client-focused; the shared issue is state and authority leaking between interactions.
Each advisory contributes its own route and prerequisites; client confirmation and database authorization are different controls.
Straiker STAR Labs asked Claude Desktop Code, an AI coding assistant, to follow an ordinary repository README. The parent could not bypass its own workspace-switch confirmation, but a Bash command launched a new headless Claude Code CLI inside the repository. That child loaded repository-local settings and ran a SessionStart hook before model inspection, sending environment data to the researcher. The demonstrated boundary failure is unmediated child-session activation; a model correctly rejecting suspicious README text did not prevent initialization code from executing.
Why it relates: The child session reaches a different trust-state path from the parent and activates the same workspace's local settings. The shared session/state boundary is workspace activation, not an established MCP authentication-session hijack.
Controlled parent-to-headless-child workspace activation and environment-data exfiltration; successful native initialization is distinguished from failed model-persuasion and directory-switch attempts.