Full Summary
This Wednesday morning, enterprises are rapidly deploying AI agents, but face significant hurdles in governance and security. JPMorgan reports a doubling of agentic AI use in large companies, jumping from 11% to 26% between 2025 and February 2026, a trend confirmed by Gartner, who predicts 40% of enterprise applications will use task-specific AI agents by year-end. However, Gartner also warns 40% of these projects could fail by 2027 due to rising costs or unclear business value. Both Gartner and ITWeb highlight a critical issue: many organizations lack proper oversight. ITWeb reveals 88% of organizations reported an AI agent security incident in the last year, with 41% to 44% lacking human control over high-risk decisions. Gartner recommends a "proportional governance" approach, classifying agents by autonomy levels to prevent over-restriction or dangerous under-restriction. In finance, QA Financial and Fox Business confirm banks are experimenting with autonomous agents for fraud and compliance, while Robinhood, as reported by Fox Business, The Verge, and Bitcoin News, is launching new agentic AI tools for stock trading. Robinhood allows users to connect third-party AI agents, like those built on ChatGPT, to manage dedicated brokerage accounts, though the company warns of significant risks including total loss of investment. On the security front, Tenable has launched Hexa AI, an advanced AI engine for cybersecurity, designed to orchestrate workflows and neutralize risk. Meanwhile, Dell and Microsoft are pushing an "agentic AI PC strategy," moving AI workloads from expensive cloud inference to on-device processing to cut rising cloud token costs, a development SiliconANGLE confirms. This means the AI systems making decisions and even trading stocks could soon be operating from your own computer, but without proper safeguards, these powerful tools pose substantial financial and security risks to your data and investments.