#USA #AI Governance #Vienna #Redwood Software #Ungoverned AI

The High Cost of Ungoverned AI Agents

Recent research reveals a concerning trend among businesses worldwide: a significant majority are encountering adverse effects from ungoverned AI agents. According to a report by Redwood Software, 77% of enterprises have suffered because of their failure to effectively manage AI's deployment and governance.

The Rapid Expansion of AI Adoption

As industries increasingly embrace artificial intelligence, they are often not prepared for the complexity of integrating these autonomous agents. The report, titled "Enterprise Benchmarks in AI and Agentic Orchestration 2026", surveyed 300 executives and specialists in technology, risk, and AI. It highlights the disparity between rapid AI adoption rates and the development of necessary governance structures.

Despite the growing recognition of AI’s potential, many organizations face challenges in maintaining control over these systems. Unexpected costs related to cloud services, operational disruptions, and compliance issues are among the most reported difficulties stemming from ungoverned AI.

Key Findings: A Call to Action

The report outlines several critical findings:

• - Financial Impact: The most common issue faced by enterprises is unexpected cloud costs, which reflects ineffective resource management linked to AI agents. Operational disruptions and compliance issues follow closely behind.
• - Scaling Challenges: While 89% of organizations increased their deployment of agentic AI in the past year, only 42% report having robust mechanisms to connect these agents with existing legacy systems.
• - Fragmented Responsibility: Responsibility for managing agentic AI is diffusely shared among different teams, including IT operations and AI-focused groups, which hampers accountability and governance.

Only a small fraction of respondents—54%—claimed to have the infrastructure necessary to effectively orchestrate AI across their business processes. Additionally, just 51% indicated they have real-time monitoring in place to observe the behavior of AI agents.

Insights from Executives

Industry leaders emphasize the urgency of addressing these issues. Charles Crouchman, Chief Product Officer at Redwood, states, "The demo might be straightforward, but the real challenge arises when an AI agent interacts directly with production systems or customer transactions. Governance shouldn’t be a post-incident consideration but integrated into the operational processes."

To effectively manage AI deployments, 62% of survey respondents noted that enhancing visibility into AI operations would significantly help. Furthermore, 58% believe proven integration capabilities with legacy systems are essential for expanding AI usage.

The Path Forward

As organizations race to leverage the power of AI, it’s vital to build robust governance frameworks around these technologies. Without this, companies are not just risking financial consequences but also delving into uncharted territory where AI-driven decisions can lead to unforeseen repercussions.

Moving forward, enterprises must prioritize the creation of centralized registries for AI agents and invest in monitoring technologies that provide real-time insights. Understanding how agents interact with both legacy systems and cloud applications will be key to mitigating risks associated with their deployment.

In conclusion, while the potential benefits of AI are immense, organizations must address the operational challenges stemming from ungoverned agents. Creating cohesive, accountable frameworks will not only help businesses navigate the complexities of AI but also unlock its transformative potential effectively.

About Redwood Software

Redwood Software stands as a pioneer in orchestration platforms focused on enabling autonomous enterprise operations. With over 30 years of expertise, Redwood assists businesses in streamlining processes and integrating AI sustainably and effectively. For more information, visit their website.