As the AI investment supercycle unfolds, Google Cloud leaders pin AI success to data, trust and change management
Artificial intelligence (AI) is no longer a future capability to be explored at the margins of the enterprise. It is becoming the defining platform shift of this decade and is reshaping how organizations operate, compete and create value across industries.
In an interview during RCR Tech’s recent AI Infrastructure Week virtual event (available on demand), Google Cloud AI Practice Leader Guruaj Bhat said, “We are frankly living through the biggest platform shift since the internet.” The implication, he said, expands beyond adoption of new technologies into rethinking business models, operating structures and the relationship between people and increasingly capable AI systems.
Yet enterprises face a familiar tension: how to move quickly with AI to capture value, while also building for the long term. Bhat describes this as an “innovation paradox” that Google Cloud sees repeatedly across customer engagements. “In the short-term you cannot afford analysis paralysis,” he said. Enterprises need to identify “high-impact, low-risk pilots to solve a specific business problem or friction point.” These early initiatives “will prove value quickly, generate excitement and build momentum.”
However, Bhat warned, “if you do only pilots” organizations risk creating “a scattered landscape of solutions” that becomes difficult to scale or govern. The solution is to operate on two parallel tracks. “While you build these short-term initiatives, you focus simultaneously on building a very unified AI platform with the relevant security and governance controls.” He compared the approach to building a city — you can open a shop quickly, but you must also build the underlying infrastructure needed to support future skyscrapers. “Don’t wait for the foundation to be perfect to start building.”
Data for AI, AI for data
Laying that foundation and building on top of it, however, depends heavily on data readiness. AI projects often expose long-standing data challenges around silos, quality and governance, particularly in large enterprises.
“AI is as good as data…it’s not only to ensure that AI provides the right outcomes but also for data you have to monetize the data, you require AI. They’re like the twins that go hand in hand,” Bhat said. Rather than waiting years to “fix” data before deploying AI, Google Cloud is encouraging customers to reverse the logic. “We don’t want perfection to be the enemy of good,” Bhat said. “What we are suggesting to customers is you don’t need to spend multiple years in cleaning the data before you start using AI. In fact, AI is the best tool to fix your data problems.”
Using tools such as Gemini to tag unstructured data, identify duplicates and enrich metadata, organizations can accelerate data maturity while improving business outcomes. “We are able to accelerate the data maturity in the organization…while delivering the business value.”
Humans in the loop delivering a “culture of curation”
If data is the foundation, trust is the differentiator as enterprises adopt agentic AI systems that act increasingly autonomously within defined guardrails. “Getting the technology right is actually the easy part,” Bhat said. “The hard part is…the cultural operating model.” Agentic AI, he added, “is going to reshape every industry, every role type, every task that is out there.”
That transformation must be led from the top. Leaders, Bhat said, need to shift the narrative from “AI will replace you” to “AI will promote you,” as organizations move “from a culture of creation to a culture of curation.” Trust is built through transparency and human-in-the-loop design. “Employees will never trust a black box agent,” he said. Agents should cite sources, link back to policy documents and explicitly say when they do not know an answer.
These themes are playing out clearly in telecom, where AI adoption must scale across complex operational environments. Google Cloud Consulting Head of TME Adeel Khan works closely with communications service providers, including Vodafone, to turn AI ambition into production reality.
“Our work is focused on tangible, high-impact business outcomes,” Khan said. At Vodafone, Google Cloud embedded a generative AI team directly into the organization to drive rapid deployment of priority use cases. The results included operational systems such as an assistant for responding to RFPs with highly accurate, tailored answers, and HR assistants designed to respond empathetically to employee queries. “These are key to realizing triple X millions of business benefits across those programs,” Khan said.
Crucially, this was not a series of disconnected experiments. “On the ground we have a team, stateful, with the customer to do a lot of that detailed work,” he said, providing a single front door to identify, prioritize and industrialize AI use cases.
From pilots to production and point solutions to platforms
For large enterprises, Khan emphasized, moving from proof of concept to production remains the biggest hurdle. “There is no silver bullet,” he said. “The transition from [proof of concept] to production is the biggest challenge for any large enterprise.”
The answer lies in platformization. At Vodafone, AI initiatives are built on Google’s Vertex AI Platform. “That platform really is the front door for all of the things we’re doing at Voda[fone] and many other customers,” Khan said, ensuring standardization, speed and responsible AI guardrails.
Equally important is change management. “Success is not just technology but the organizational change management that goes alongside that,” he said. Google Cloud works with customers on learning, enablement and business transformation plans to build confidence at every level of an organization.
Adaptability is the new competitive advantage
Looking ahead, both executives see AI journeys unfolding in phases: near-term efficiency gains, followed by growth and new revenue models. In telecom, that includes AI-driven customer insights, network lifecycle automation and, longer term, new AI-enabled services and marketplaces.
What is not optional, Khan warned, is inaction. “Inertia I don’t think is acceptable…You will get left behind and that could be an existential question…It’s a threat if you don’t do anything. It’s an opportunity to get ahead of the game.”
Bhat echoed the urgency. “We are in the midst of this platform shift,” he said, noting that the pace of change is faster than ever. His advice is to “start now and stay flexible…The companies that will lead their industries in five years are the ones who will get started quickly, who will ground the models with their data, upskill their workforce.”
The cost of waiting, he concluded, may be higher than the cost of experimentation. “The cost of not acting will be a significant challenge for these companies if they do not accelerate the innovation with AI right here, right now.”
