AI Sec Watch

STADLER, a 230-year-old recycling equipment company, embedded ChatGPT (an AI language model that generates human-like text) across its workforce to speed up knowledge work like drafting, summarizing, and translating. The company achieved 30-40% time savings on common tasks, 2.5x faster first drafts, and 85% daily active usage by providing company-wide access, training, and clear guardrails while encouraging bottom-up experimentation.

Langflow, a tool for building AI-powered agents and workflows, had a vulnerability in versions before 1.9.0 where the Agentic Assistant feature would execute Python code generated by an LLM (large language model) on the server. An attacker who could access this feature and control what the model outputs could run arbitrary code (malicious commands) on the server itself.

Nanobot, a personal AI assistant, had a vulnerability in its email module that allowed attackers to send malicious prompts via email, which the bot would automatically process as trusted commands without the owner's knowledge. This is a type of indirect prompt injection (tricking an AI by hiding instructions in its input) that could let attackers run arbitrary system tools through the bot. Version 0.1.6 fixes this flaw.

LibreChat versions 0.8.2-rc1 through 0.8.3-rc1 have a vulnerability where user-created MCP (Model Context Protocol, a system for connecting AI models to external tools) servers can steal OAuth tokens (security credentials used for authentication). An attacker can create a malicious MCP server with special headers that trick LibreChat into substituting sensitive tokens, which are then leaked when victims use tools on that server.

LibreChat (a ChatGPT alternative with extra features) versions 0.8.2-rc2 through 0.8.2-rc3 have a security flaw in the SSE streaming endpoint (a real-time data connection) at `/api/agents/chat/stream/:streamId` that fails to check if a user actually owns a chat stream. This means any logged-in user can guess or obtain another user's stream ID and read their live conversations, including messages and AI responses, without permission.

LibreChat (a ChatGPT alternative with extra features) versions 0.8.2-rc2 through 0.8.2 have a vulnerability that allows attackers to access internal systems through SSRF (server-side request forgery, where an attacker tricks a server into making requests to resources it shouldn't access). Even though a previous SSRF fix was applied, it only checked domain names and didn't verify whether those names actually point to private IP addresses (internal network addresses), leaving the system exposed.

LibreChat, a ChatGPT alternative with extra features, has a security flaw in versions before 0.8.3 where a function called `isPrivateIP()` fails to recognize IPv4-mapped IPv6 addresses (IPv6 addresses that contain IPv4 address information) in a certain format, allowing logged-in users to bypass SSRF protection (SSRF is server-side request forgery, where an attacker tricks a server into making requests to internal networks it shouldn't access). This could let attackers access sensitive internal resources like cloud metadata services and private networks.

LangChain Core has a path traversal vulnerability (a security flaw where attackers can access files outside intended directories using '../' sequences or absolute paths) in legacy functions that load prompt configurations from files. When an application accepts user-influenced prompt configs and passes them to `load_prompt()` or `load_prompt_from_config()`, attackers can read arbitrary files like secret credentials or configuration files, though they're limited to specific file types (.txt, .json, .yaml).

Langflow had a vulnerability where the code checking if a user owned a flow was missing when authentication was enabled, allowing any authenticated user to read, modify, or delete flows belonging to other users, including stealing embedded API keys. The fix removes the conditional logic and always checks that the requesting user owns the flow before allowing any operation.