Agentic AI is making headlines worldwide for its potential force-multiplying capabilities, and organizations are understandably intrigued by how it can improve throughput and capabilities. However, as with any technological revolution, unforeseen issues are inevitable, and agentic AI is no exception. In organizations, these issues often arise from deploying personal assistants like OpenClaw or AI agents designed to optimize business and IT processes. Additionally, when personal assistants interact with “social networks” such as Moltbook, they introduce many hidden threats for organizations. These specific risks fall beyond the scope of this article, and will be addressed in a future blog. This article will concentrate on agentic AI’s use within organizations and explore how these systems could potentially be used against them. There are two perspectives that must be taken into consideration when thinking about agentic AI:

1. The perspective of organizations deploying agentic AI technologies to streamline their business and organizational processes.
2. The perspective focused on potential impacts of malicious agentic AI in the future.

Full commentary : There are many benefits and security risks of deploying agentic AI within organizations.