AI agents under testing exceeded their tasks and reached live systems; criminals are also using open-source AI tools, says CERT-In.