ServiceNow's Sumeet Mathur says rising tokenisation costs are a 'good driver' for enterprises to use AI more efficiently, as adoption shifts from POCs to real deployments.
Alpha Score of 34 reflects weak overall profile with weak momentum, poor value, strong quality. Based on 3 of 4 signals – score is capped at 90 until remaining data ingests.
Sumeet Mathur, Senior Vice President and Managing Director of ServiceNow's India Technology and Business Center, argued that the rising cost of tokenisation is a positive force for enterprise AI adoption. The price pressure, he said, forces companies to use models more efficiently rather than discouraging adoption.
"To me, tokenisation is a good driver for enterprises to ensure they are using AI in a proper way," Mathur said in an interview. "Cost as a constraint actually drives the right behaviours."
The comments come as enterprise AI spending shifts from proof-of-concept to real deployment. Mathur said that until a few months ago, the company saw many demos and POCs in its client base. Now, enterprises are adopting AI in earnest and asking ServiceNow for help because initial deployments are not generating the value they expected.
"We are seeing a lot of traction to route, escalate, automate and resolve security threats, and it is showing in our numbers," he said. ServiceNow has committed to the street that it will generate $1.5 billion in revenue from AI-related products by the end of next year.
Mathur said the industry is realising that context engineering is more important than the large language models themselves. "The efficacy is determined on the context you provide, instructions, memory, knowledge graph, security guardrails and governance," he said. "Context is more important in enterprise AI than just models and prompts."
On token costs, Mathur pushed back against the notion that high prices are a barrier. "People are looking at it and saying, why are my software developers running out of tokens? How are they using it? Are they using it properly or not?" he said. "You don't always have to use the most expensive model." The constraint, he argued, will lead to better optimisation because companies will think harder about how they use AI.
When asked about data sovereignty, Mathur expanded the definition beyond physical location. "It is about ensuring that right data gets used with the right access control in the right context in the flow of work and operations that my enterprise is doing," he said. ServiceNow works with all major data lake and cloud providers and does not restrict which models or hyperscalers its customers use.
"Cost as a constraint will lead to better work with AI because we will think about how to optimise it," Mathur said.
Drafted by a large language model from the source reporting linked above, then screened by automated publishing checks. It is not read by a journalist before publication. Some articles cite our Alpha Score. Verify prices and figures against the original source. Educational coverage, not personalized advice.