GenAI is disrupting industries across healthcare by revolutionising medical outcomes for patient finance by enhancing automation and process refactoring , education by facilitating adaptive learning but Agentic AI is very hard to move from POC/Pilot to production scale .

US companies have invested between $35 and $40 billion in Generative AI initiatives and, so far, have almost nothing to show for it.

According to a report from MIT's NANDA (Networked Agents and Decentralized AI) initiative, 95 percent of enterprise organizations have gotten zero return from their AI efforts. Only 5 percent of organizations have successfully integrated AI tools into production at scale. The report is based on 52 structured interviews with enterprise leaders and on analysis of more than 300 public AI initiatives and announcements, and a survey of 153 business professionals.

The report authors – Aditya Challapally, Chris Pease, Ramesh Raskar, and Pradyumna Chari – attribute this GenAI Divide not to insufficient infrastructure, learning, or talent, but to the inability of AI systems to retain data, to adapt, and to learn over time.

The GenAI Divide is starkest in deployment rates, only 5 percent of custom enterprise AI tools reach production

The report says "Chatbots succeed because they're easy to try and flexible, but fail in critical workflows due to lack of memory and customization

While about 50 percent of AI budgets get allocated to marketing and sales, the report authors suggest that corporate investment instead should flow toward activities generating meaningful business results. This includes lead qualification and customer retention on the front end and, in the elimination of business process outsourcing, ad agency spending, and financial service risk checking on the back end. 

Companies that bridge the GenAI divide approach AI procurement as business process outsourcing customers rather than as software-as-a-service clients, the authors argue.

They demand deep customization, drive adoption from the front lines, and hold vendors accountable to business metrics," the report concludes. "The most successful buyers understand that crossing the divide requires partnership, not just purchase
The current lack of enterprise-ready guardrails at scale to understand and manage risks, built into generative AI systems is a key concern for organisations. 

As the technology is still in its early stages, many organizations are worried about not having the right tools, expertise, or processes to effectively manage and mitigate the risks associated with using Agentic  AI. While businesses know that adopting AI is critical to remain competitive, there is a low-risk appetite for AI technologies which may damage company reputations ,business risk,compliance frameworks and testing frameworks.

An Example: A financial services institution uses an Agentic AI system to provide investment recommendations to their traders. The system generates a convincing report suggesting a specific stock, but the report is a hallucination based on training data, not on the stock's actual performance. A client invests based on the misleading report and suffers significant financial losses when the stock underperforms. This case could lead to reputational harm, financial losses, legal liabilities, and compliance issues for the financial institution. 

In Summary the problem is complex as for  Agentic AI to be  successfully deploying this needs the right integrations that are secure , access to the appropriate data that has good governance and provenance  and robust novel Pen testing/Jailbreak/Redteam/Context engineering frameworks that can handle non deterministic AI systems who have a tendency to Hallucinate, propagate Bias, Leak PII, Ignore data privacy laws ,spread misinformation ,harmful content and cause cybersecurity risks.