Companies Are Building AI Workforces, but Many Still Need to Deploy Governance Systems
When AI agents join the workforce a new toolset is needed to manage them.
New research shows agentic AI use growing more than fivefold as workers increasingly manage multiple agents and delegate day-long assignments
Companies would never give an employee broad access to corporate systems, provide no clear responsibilities, and hope an audit log explains the consequences afterward, but they're doing that with AI.”
SAN DIEGO, CA, UNITED STATES, July 14, 2026 /EINPresswire.com/ -- Companies are rapidly adding a new kind of worker to their organizations.— Adam Harriss
It can access company data and business systems. It can perform complex assignments. It may work continuously and take actions on behalf of employees.
But in many organizations, no one has clearly defined its job, who manages it, how much authority it has, or how its access should be revoked.
That worker is an AI agent.
DoubleU.ai, Inc., the company behind DBLU, says businesses need to begin treating AI-agent management as an organizational discipline, not simply a software deployment problem. The warning follows new research, The Shift to Agentic AI: Evidence from Codex, which found that active users of OpenAI’s Codex increased more than fivefold during the first half of 2026.
The study also found that more than 10% of users manage three or more concurrent agents during a typical week. Since the beginning of 2026, the share of users assigning at least one task estimated to require more than eight hours of experienced human work increased nearly tenfold.
People are no longer simply asking AI questions.
They are delegating work to it.
“Companies are beginning to manage AI workers before they have built a management system for AI,” said Adam Harriss, Founder and CEO of DoubleU. “Every meaningful AI agent should have the equivalent of a job description, an identity, a manager, a defined purpose, an authority level, approval limits, and a record of what it did.”
Organizations have spent decades building systems for human authority. Employees have identities, roles, managers, permissions, policies, approval limits, and onboarding and offboarding processes.
AI agents increasingly perform work, but many businesses cannot answer equivalent questions:
- Who is the agent acting for?
- What job has it been assigned?
- What data and systems can it access?
- What decisions can it make?
- What requires human approval?
- Who is accountable when it makes a mistake?
- How is its authority revoked?
According to IBM Institute for Business Value research conducted with Oxford Economics, “87% of surveyed executives say they have clear AI governance frameworks, but fewer than 25% have fully implemented and continuously review tools to manage risks such as bias, transparency, and security.”
Corroborating those findings, recent DigiCert research found that while 90% of organizations have discussed AI governance at the executive or board level, only half have established formal AI governance programs. The same research found that only 53% can fully trace AI outputs back to the underlying models and source data.
“Companies would never hire an employee, give that person broad access to corporate systems, provide no clear responsibilities, and hope an audit log explains the consequences afterward,” Harriss said. “Yet that is remarkably close to how many organizations are beginning to deploy AI.”
DBLU is building a Trust Layer designed to operate at runtime and help organizations manage AI identity, purpose, permissions, delegated authority, human approvals, governed memory, and auditability.
The company argues that enterprise software was built around a simple assumption: a human was behind every important action. AI agents break that assumption.
“The next management challenge is not simply how to use AI,” Harriss said. “It is how to manage a workforce that is no longer entirely human.”
Adam Harriss is available for interviews on the emergence of AI workers, the risks of unmanaged AI agents, and why companies may need a new management model for non-human labor.
The research paper, The Shift to Agentic AI: Evidence from Codex, is available at [https://arxiv.org/html/2606.26959v1].
The IBM study, IBM Study: AI Spending Expected to Surge 52% Beyond IT Budgets as Retail Brands Embrace Enterprise-Wide Innovation, is available at [https://newsroom.ibm.com/2025-01-07-ibm-study-ai-spending-expected-to-surge-52-beyond-it-budgets-as-retail-brands-embrace-enterprise-wide-innovation]
The Digicert research report, AI Trust Outlook: AI is scaling faster than trust, is available at [https://www.digicert.com/content/dam/digicert/pdfs/report/ai-trust-pulse.pdf]
About DoubleU
DoubleU.ai, Inc. is building DBLU, a Trust Layer for AI agents. DBLU helps organizations govern who AI is acting for, what information it can use, for what purpose, what authority it has, when human approval is required, and how its actions can be audited.
For interviews or more information:
Adam Harriss
Co-Founder & CEO, DoubleU.ai, Inc.
Email: press@doubleu.ai
Phone: 858.449.9030
Website: https://dblu.ai
Adam Harriss
DoubleU.ai, Inc.
+1 858-449-9030
email us here
Visit us on social media:
LinkedIn
Facebook
Legal Disclaimer:
EIN Presswire provides this news content "as is" without warranty of any kind. We do not accept any responsibility or liability for the accuracy, content, images, videos, licenses, completeness, legality, or reliability of the information contained in this article. If you have any complaints or copyright issues related to this article, kindly contact the author above.
