The 18-Fold Cost Explosion
Recent findings from technology research firm Gartner have sent a ripple of concern through the business world. According to reports from the Gartner IT Symposium/Xpo 2026 in Kochi, the investment required to generate positive returns from AI has skyrocketed
by 18 times in just two years. This 'AI inflation' highlights a growing disconnect between the initial excitement for AI and the practical reality of its commercial payoff. While over half of business leaders report some positive outcomes from AI, the soaring cost to achieve those gains suggests that early wins are not translating into sustainable, cost-effective strategies. The issue isn't just about the price of software or hardware; it encompasses a wider ecosystem of costs, including implementation, infrastructure, and crucial ongoing oversight.
The Rise of Autonomous Agents
Compounding the cost issue is the rapid move toward the next wave of AI: autonomous agents. These are not just chatbots or simple automation tools. AI agents are systems designed to take action, execute multi-step tasks, and operate with a degree of independence. Gartner's research indicates a huge appetite for this technology, with 71% of organisations expecting to deploy agentic AI within the next year. However, this rush to adopt more powerful, autonomous systems creates a significant challenge. As AI evolves from a simple assistant to an active participant in workflows, the potential for unintended consequences, security risks, and runaway costs multiplies. A single request can trigger multiple model calls and tool interactions, leading to non-linear cost expansion that many businesses are unprepared for.
A Widening Governance Gap
Here lies the core of Gartner's warning: while the majority of companies are planning to deploy AI agents, a mere 21% currently have the governance capabilities to oversee them. This creates a massive 50-point gap between ambition and readiness. Without proper governance, businesses are flying blind. This lack of oversight is a primary reason Gartner predicts that over 40% of agentic AI projects will be canceled or decommissioned by 2027 due to post-production incidents. Ungoverned agents can expose sensitive data, make poor decisions, or enter costly loops that burn through budgets without delivering value. Incidents of AI-generated 'work slop' are already costing large companies millions in wasted time as employees correct errors.
Introducing AI TRiSM
Gartner's answer to this challenge is a framework called AI TRiSM, which stands for AI Trust, Risk, and Security Management. This isn't a single product, but a discipline and a set of practices designed to turn abstract governance policies into tangible, operational controls. AI TRiSM provides a structured approach for managing AI systems throughout their entire lifecycle. It focuses on continuous monitoring and real-time enforcement rather than static, outdated policy documents. The framework is built on pillars including explainability (understanding why an AI did what it did), ModelOps (managing the lifecycle of models), AI application security, and data privacy. By operationalizing governance, AI TRiSM allows organisations to expand their use of AI confidently, knowing that guardrails are in place to manage risks and costs.
The Way Forward for Indian Businesses
For Indian companies enthusiastically adopting AI, this warning is particularly timely. The temptation is to move fast to capture a competitive edge, but scaling too quickly without a governance strategy is a recipe for failure. As Gartner analyst Neha Kumar noted, organisations need to be more selective about the projects they pursue. The focus must shift from simply asking if AI can perform a task to rigorously evaluating if the productivity gains justify the total cost of ownership. This means implementing robust governance from day one, tracking costs granularly, and ensuring every AI initiative is tied to clear business value. The goal is to put brakes on the car not to stop it, but as Gartner's Daryl Plummer says, to allow it 'to go fast safely'.
















