Companies Are Giving AI More Authority Than Accountability

As AI moves from generating answers to taking actions inside business workflows, JOTO PR points to a widening gap between what companies allow AI to do and the operating models, boundaries, and controls needed to govern it.

TAMPA, Fla., Sept. 10, 2026 /PRNewswire/ -- Artificial intelligence (AI) is taking on more authority inside businesses workflows while many organizations are still defining the rules for how that authority should work. Deloitte found that only 5% of organizations say their business processes are highly prepared for AI agents, even as 61% of executives expect most agents they use to become generally autonomous within four years.

Karla Jo Helms, Chief Evangelist and Anti-PR® Strategist at JOTO PR Disruptors™, says the findings point to the next phase of the enterprise AI problem: companies are no longer deciding whether employees should use AI. They are deciding how much authority software should have to act on their behalf. "Companies spent the first phase of AI asking what the technology can do," said Helms. "The next phase is about authority: what should AI be allowed to do, who decides, how those limits are enforced, and who owns the outcome when the system acts."

Production AI Needs an Operating Model

The first accountability problem appears when AI moves from experimentation into everyday business processes.

Ben Prescott, Head of AI Solutions at Trace3, a leading IT and AI solutions provider, says production changes the equation because AI begins affecting more users, workflows, and business outcomes.

"The biggest gap here is not rolling out a production app," Prescott said. "It's getting people to use it, learn how to trust it, or how to measure it." Those challenges become more relevant as AI begins performing work rather than simply supporting it. A system can function exactly as designed and still fail if no one owns the outcome, employees do not use it as intended, or the company cannot show that it is improving the workflow it was intended to support.

Autonomy Does Mean All or Nothing

The second problem is how much authority an AI system should receive.

SHIP.com, an AI-powered shipping platform designed to help ecommerce businesses, provides a practical example through shipping, where an automated action can create an immediate financial consequence. Its system allows users to establish automation rules, including requiring approval before an order ships or flagging actions above a defined spending level. Sellers can choose how proactive they want the system to become rather than accepting a single level of control.

"Are you going to give the employee the credit card and say, 'Go do whatever you want,' or do you want to have some authority over what they're spending it on?" said Joe DiSorbo, Founder and CEO of SHIP.com.

The principle extends beyond shipping. AI may be able to send communications, change records, trigger workflows, purchase services, and act across connected systems. The fact that a system can perform an action does not determine whether it should have unrestricted authority to do so.

Trust Versus Control: Policies Need Enforcement

The third problem appears after companies establish rules.

IBM's 2025 Cost of a Data Breach research found that 97% of organizations reporting an AI-related security incident lacked proper AI access controls. One in five organizations also reported a breach connected to shadow AI, showing how quickly formal policies can lose relevance when actual use is not visible or controlled.

But AI governance doesn't end with written policies. "A policy without enforcement is simply an expectation," said Justin Cameron, Chief Technology Officer of Magna5, a national managed IT, cybersecurity, cloud, and compliance services provider. "Organizations need to take a two-pronged approach by giving employees access to a secure, approved AI platform, and maintaining visibility into whether unsanctioned tools are being used outside of it. Without both in place, companies are relying on trust rather than control."

The Cost of Getting Autonomy Wrong

The consequences of weak governance may soon become visible in production. Gartner predicts that by 2027, 40% of enterprises will demote or decommission autonomous AI agents because of governance gaps identified after production incidents. For Helms, that distinction captures the larger accountability problem. As AI takes on more consequential work, authority should expand alongside the operating models, boundaries, monitoring, and enforcement needed to control it.

"Giving AI more autonomy does not transfer accountability away from the business," Helms said. "The companies that scale AI successfully will not be the ones that give agents the most freedom. They will be the ones that develop the guidelines for where that freedom starts, where it stops, and who is responsible for what happens in between."

About JOTO PR Disruptors™
JOTO PR Disruptors™ is the Anti-PR® agency behind some of the fastest-growing tech and innovation brands in the U.S. By applying crisis management techniques and modern media algorithms, JOTO PR creates third-party credibility campaigns that produce measurable results and build undeniable market influence. Learn more at www.jotopr.com. 

References

• Deloitte. (2026, August 12). AI agents are only the beginning: Deloitte survey examines the AI readiness gap and reveals how enterprises can prepare for agentic success. Deloitte. deloitte.com/us/en/about/press-room/deloitte-survey-examines-ai-readiness-agentic-ai-success.html
• Gartner, Inc. (2026, May 26). Gartner says applying uniform governance across AI agents will lead to enterprise AI agent failure. gartner.com/en/newsroom/press-releases/2026-05-26-gartner-says-applying-uniform-governance-across-ai-agents-will-lead-to-enterprise-ai-agent-failure
• IBM. (2025, July 30). IBM report: 13% of organizations reported breaches of AI models or applications, 97% of which reported lacking proper AI access controls. IBM Newsroom. newsroom.ibm.com/2025-07-30-ibm-report-13-of-organizations-reported-breaches-of-ai-models-or-applications%2C-97-of-which-reported-lacking-proper-ai-access-controls

Media Contact:
Karla Jo Helms
JOTO PR™
727-777-4629
jotopr.com 

SOURCE JOTO PR