AI Risks · SECURITY.COM

Sandbox Flipped to ' Danger -Full-Access': Unauthorized AI Relays Create Risk for Organizations

A grey market relay service installs the real OpenAI Codex and Anthropic Claude Code CLIs, then reconfigures them to route every prompt and ...

Read the original at SECURITY.COM ↗

Mentions
AnthropicClaudeOpenAI
Tracked via
“AI Risks”

Same story, other outlets