Researchers Find 175,000 Publicly Exposed Ollama AI Servers Across 130 Countries

A joint SentinelLabs and Censys investigation found 175,000 publicly exposed Ollama hosts across 130 countries—nearly half advertising tool-calling capabilities—creating high-severity risks when unauthenticated and internet-facing, including prompt injection and abuse of victim infrastructure. The report also outlines an active LLMjacking campaign, Operation Bizarre Bazaar, attributed to “Hecker,” which scans for exposed Ollama/vLLM/OpenAI-compatible endpoints, validates quality, and resells discounted access via silver.inc, underscoring governance gaps across cloud and residential deployments and urging strong authentication, monitoring, and network controls for edge AI services.