“AI is changing the world and we are leading the charge,” Google proclaimed at its recent cloud summit in London. Google can reasonably claim to be a pioneer of generative AI technology with Gemini, and though it wasn’t the first to make waves in the market, it’s determined to make up for lost time. The huge amount of effort and resources Google is putting into its AI strategy and building out an extensive ‟full stack” portfolio—at the model level with Gemini, at the machine learning level with Vertex AI, at the infrastructure layer, and at the data analytics layer with BigQuery—would appear to validate the scale of its intent here to establish itself at the forefront of the new AI revolution.

To put some numbers around the scale and breadth of its efforts here, Google has delivered over one thousand individual product enhancements across Google Cloud and Workspace over the last 12 months. And Google Cloud’s software business has a $40 billion run rate, making it the world’s fourth-largest software company.

AI has now become the central and differentiating point of Google Cloud overall. It’s leveraging its heritage as a company with unrivaled technology to distinguish itself from the other major public cloud providers, which it paints as more infrastructure-oriented. Customers are placing increased value on providers that can offer the complete tech stack, and Google says its experience in delivering this is enabling it to onboard new customers more quickly and deliver AI at scale as the technology matures.

This perspective was validated by the many large enterprise organizations presenting at the Google Cloud Summit, many of whom highlighted substantial savings and other benefits that their AI investments were already delivering. Lloyds Bank is using Google Cloud AI across a range of use cases, including code translation, threat intelligence, and document categorization and summarization—the latter of which had alone delivered £15 million in cost savings. Another bank, HSBC, is using Google Cloud AI in its financial crime division, and as a result has seen its case volume lowered 70% by significantly reducing the volume of noise. This has enabled HSBC to increase the amount of crime it identified by up to four-fold, letting its case handlers focus on more complex crimes. And multiple retailers, including Kingfisher (and its subsidiary Screwfix) and Dunelm demonstrated their use of Vertex AI-powered image search to boost the effectiveness of their recommendation engines. All customers cited Google’s reputation as an innovator in AI as a key factor in selecting it.

However, an exceptional product portfolio doesn’t necessarily translate into market success and leadership. Recognizing that having a large and ever-expanding portfolio of capabilities runs the risk of becoming unwieldy for customers, a big part of Google’s focus moving forward is to bring these capabilities together to deliver more integrated experiences for customers. Indeed, by focusing on outcomes rather than merely actions Google believes it can help bring its platforms to life.

One part of this strategy is to invest in partnerships, supporting Google’s commitment around delivering optionality for customers. This is a broad effort that ranges from the model side, with support for third-party models and open source (e.g. Meta/Llama), to offering support for third-party storage capabilities within Google Cloud (e.g. Google Cloud NetApp Volumes) to building partnerships with hardware providers that can help enable Google Distributed Cloud for on-premises cloud deployments.

Another strand is to continue investing in region- and industry-specific capabilities. Here, Google Cloud announced additional data residency commitments, including the ability for UK-based organizations to run machine learning processing for Gemini 1.5 Flash within the UK, helping UK-based organizations, including public sector, address data sovereignty and compliance concerns. This is on top of Google Cloud’s previously announced $1 billion investment in a new UK data center, and range of investments in the UK startup ecosystem.

A final focus area here is continuing to invest in more integrated product capabilities that take advantage of Google’s broad portfolio. Unsurprisingly, a major focus area here is in the fast-evolving worlds of intelligent agents, such as Agent Builder, part of the Vertex AI platform. This is where Google believes it can bring to bear its entire stack of expertise to synthesize across multiple models (speech, text, audio, video, etc.), in a natural manner and delivered almost instantly using its heritage in search. It’s early in its deployment, but the company is investing in developing capabilities in several specific areas, including customer service, creative functions, data, code development, and security.

It notes interest across the board, here with particular interest in the marketing realm. Of the Google customers presenting at the summit, few had reached implementation phase, but many were examining the potential of deploying large-scale agentic AI. “The use cases are compelling; the question is whether humans will trust it,” noted one customer, adding “But we’re going to try very hard to make it work.”