Inside HSBC’s Multi-Year AI Pact With Google Cloud: 200+ Use Cases, $100M+ Per Project

Breaking down the three target areas

The partnership zeroes in on three specific domains:

• Hyper-personalised wealth management support: AI will generate tailored advice and product recommendations for clients, allowing the bank to offer services typically reserved for high-touch advisory relationships at scale .
• Stronger financial crime risk management: Enhanced machine learning models will detect money laundering and other illicit activity with greater accuracy, building on HSBC’s existing AI-based transaction monitoring infrastructure .
• AI tools for frontline and relationship managers: Staff will receive decision-support capabilities that improve client service, speed up workflows, and reduce manual processes. These tools are designed to augment human judgment, not replace it entirely .

The tech stack: Gemini, DeepMind, and agentic AI

The technical backbone of the partnership draws on Google’s most advanced AI capabilities. HSBC will gain access to the latest Gemini models and the Gemini Enterprise Agent Platform, a system that enables the bank to build and deploy custom AI agents capable of automating complex, multi-step tasks .

Engineers from both Google Cloud and Google DeepMind will work directly with HSBC teams to develop these tools, making the collaboration a joint R&D effort rather than a simple vendor arrangement . HSBC’s existing cloud footprint—over 600 applications already hosted on Google Cloud—provides the infrastructure layer for this expansion .

A proven track record: the Dynamic Risk Assessment story

This is not HSBC’s first major AI collaboration with Google. The bank previously co-developed an AML system called Dynamic Risk Assessment (DRA), which uses Google Cloud’s AML AI product as its core detection engine .

DRA now monitors over one billion transactions per month and has produced measurable improvements: a two- to four-fold increase in legitimate suspicious activity detection, alongside a roughly 60% reduction in false positive alerts . Those results effectively demonstrated that AI can outperform traditional rules-based compliance systems while dramatically lowering investigation costs. The 2026 partnership applies this same co-development model to a much wider set of business problems.

What the CEO says about jobs

HSBC CEO Georges Elhedery has been blunt about the trade-offs. He has publicly stated that generative AI will destroy certain roles and create new ones, though he emphasized that people will remain central to banking even as AI usage expands .

Reports indicate the bank has considered cutting approximately 20,000 roles—roughly 10% of its global workforce—as a direct result of AI-driven efficiency gains . This aligns with broader industry sentiment: JPMorgan Chase CEO Jamie Dimon has said AI "will eliminate jobs," Citigroup CEO Jane Fraser has acknowledged some roles "will no longer be required," and Goldman Sachs President John Waldron has referred to routine banking processes as a "human assembly line" ripe for automation .

HSBC is among the first major global banks to publicly attach a hard dollar figure to its AI expectations, a move that signals how seriously leadership views the technology as a profit driver . The $100-million-per-project benchmark sets a concrete target in a sector where many AI deployments remain experimental or unmeasured.

The partnership also places HSBC at the forefront of an intensifying technology arms race among global banks. With competitors like JPMorgan, Citi, and Goldman Sachs all racing to embed AI, HSBC’s willingness to quantify the upside—and to staff accordingly—marks a significant strategic statement .