Microsoft has introduced Cyber Pulse, a digital briefing aimed at business leaders, addressing the security and governance challenges of deploying AI agents at scale. The inaugural edition highlights a shift from AI experimentation to broader integration of human and AI “agent” teams, advocating for a security framework founded on agent visibility, governance, and Zero Trust principles. Notably, a staggering 53% of Australian organizations lack specific GenAI security controls, revealing a gap in monitoring unauthorized agents. Cyber Pulse forecasts over 1.3 billion autonomous AI agents by 2028, emphasizing widespread adoption across sectors such as financial services, manufacturing, and retail. It warns of rising unsanctioned agent use, citing a 29% employee rate of utilizing unsanctioned AI tools, which complicates compliance. To mitigate risks from agent misuse, Microsoft’s framework advocates for centralized observability and control, positioning the responsibility for governance and security at the executive level.
