Agentic Contracting: How Executives Are Reconciling Urgency and Caution in the AI Era 

Introduction

Searching for ROI in the Age of Agentic AI

The honeymoon period for generative AI is ending, and the C-suite knows it. According to BCG, 74% of global companies report they are struggling to achieve or scale value from their AI programs despite heavy investment. Meanwhile, an IBM survey of 2,000 CEOs found that only about one in four AI initiatives are delivering the expected return.

And yet few companies appear to be tempering their enthusiasm for AI adoption.

A recent poll of a thousand IT and business leaders shows 51 percent already deploying agents in live workflows, with another 35 percent planning adoption by 2027. Deloitte’s 2025 technology-media-telecom forecast echoes the excitement but sounds a note of caution: true autonomy will take time, rigorous guardrails, and well-chosen use cases.

To better understand how executives are approaching these countervailing forces – enthusiasm for AI on the one hand, challenges around ROI on the other—Icertis commissioned a survey of 1,000 C-suite executives at enterprises with 5,000 or more employees in the United States, the United Kingdom, and India to drill in on one particular enterprise process receiving a massive amount of hype in the AI market: Contracting. 

With their unique marriage of natural language, financial data, and legal bearing, contracts represent a fascinating edge case for AI agent deployment: On the one hand, as revealed in this survey, contracts consume a significant amount of work hours, even in the executive ranks. This would nominate them as a prime target for AI automation. At the same time, given their foundational role in commerce, the margin for error is infinitesimal: if an AI chatbot gets a customer service interaction wrong, a company may have one angry customer to deal with; get a contract wrong and it could face huge financial, legal or regulatory consequences. 

This survey demonstrates that executives are well aware of this balancing act – and come away bullish on the value agentic contracting will deliver to their business. 

Draining Productivity

Contracting: A Hidden Drain on Executive Productivity 

The general perception of a CxO is a person focused on the big picture, not mired in the details. 

Yet as this survey shows, contracts serve as a significant exception.

More than half of respondents said they personally review and approve over 500 contracts per year, and one in five put that number above 1,000. Even more striking, 40% spend 10 hours or more a week on contracts. (Amazingly, in the age of “click to sign,” nearly 40% say they have signed a paper contract within the last year!).

These figures underscore the criticality of contracts in commerce, representing, as they do, the supplier, partner, and customer relationships that define corporate strategy.

They also underscore how resistant contracts have been to digitization—it’s likely been decades since a Fortune 500 CFO has signed a physical check for payments, yet here they are still signing the dotted line on contracts.

The reason for this is well understood: With their intricate and nuanced legal language, it’s been impossible to structure contract data the way organizations have been able to structure financial, customer, employee or supplier data. 

Yet contracts must be reviewed and signed – leaving executives no choice but to devote a significant amount of their time on them. Until now. 

Relief in AI

Finding Relief in Agentic AI

Agentic AI can seem almost tailor-made to overcome the historical hurdles to contract digitization: technology that can interpret unstructured natural language and perform specific tasks (reviewing redlines, analyzing risks) against that language.

Our survey finds that executives agree that contracts are a use case primed to deliver that elusive return on AI investment.

83% of surveyed leaders said AI agents that manage relationships with customers, suppliers, and partners would be a top-priority use case within their wider AI agenda (only 1% said they didn’t think AI applied to business relationships would have high ROI for their business). 

More specifically, 53% expect AI agents to negotiate customer or supplier deals on behalf of the business within the next 12 months. An additional 34% predict the same within one to three years. In other words, nearly nine out of ten leaders foresee agent-led negotiations becoming operational reality by 2028, and more than half believe the pivot will happen in a single budget cycle. 

Risky Business

Eyes Wide Open Approach to AI Risks

What makes this data even more revealing – to both the nature of AI as well as contracts themselves—are the risk that executives acknowledge could come with this application of AI.

• 56% say they are “very concerned” about granting agents the autonomy to make business decisions without suitable guardrails.
• Data security tops the list of perceived deployment hurdles, closely followed by compliance and reputational risk.
• 44% admit they lack sufficient trust in an agent’s ability to execute tasks autonomously.

How can leadership teams hold two seemingly contradictory views—embracing agentic contracting while fearing its autonomy? The answer lies in the urgency-risk trade-off that defines many technology transitions. Leaders recognize that contracts consume scarce executive bandwidth and that AI offers a rare structural solution. They also acknowledge that no organization can afford to be a late adopter when the payoffs include faster deal cycles, improved compliance throughput, and real-time risk insights.

At the same time, autonomy is not a binary setting. It is best conceived as a gradient. Early deployments may focus on decision support, where agents assemble relevant clauses, highlight deviations from policy, and recommend fallback language—all before a human approves the final move. Gradually, as audit trails, performance logs, and exception-handling protocols prove their reliability, organizations can extend autonomy to parallel tasks: generating first-pass redlines, cross-referencing obligations with invoices or shipping data, and even proposing remediation steps for underperforming clauses.

There is also an argument to be made that applying agents to contracts could in fact address some of the very concerns uncovered by this survey.

Contracts are essentially the “rules of business” that a company must follow in all interactions with its commercial counterparties. Think of them like the traffic rules an autonomous car must understand before it backs out of the driveway. 

Therefore, when contracts are made central to agentic workflows, they can serve as an enabler of safer—and therefore more valuable—automation across the enterprise.

Conclusion

Balancing Boldness and Prudence

The executive psyche is often caricatured as either risk-averse or disruption-hungry. The survey of 1,000 C-suite leaders paints a more sophisticated portrait—one that blends bold adoption timelines with clear-eyed vigilance. Leaders recognize that contracts remain the lifeblood of revenue and cost control; they also acknowledge that manual stewardship is no longer tenable at enterprise scale. AI agents promise a way out of that bind, but only if risks are managed with the same rigor that governs any mission-critical system. 

Ultimately, the path forward is neither a headlong rush into full autonomy nor a defensive crouch behind legacy processes. It is an incremental ascent—calibrated autonomy, layered security, continuous oversight—anchored in the conviction that well-designed agentic systems can elevate human judgment rather than replace it. In that future, executives will reclaim thousands of hours once lost to clause revisions and compliance spot-checks, redeploying their attention to innovation, growth, and societal impact. The survey signals that the journey has already begun; the next 12 months will reveal which organizations turn insight into sustained competitive advantage.