Financial Services Has Entered AI’s Next Phase: Measurable, Governed and Agentic

Executive Summary

• Financial services companies have rapidly progressed in AI use, shifting from AI pilots and limited use cases to accountability. Firms are increasingly focused on measurable business value, with 68% of financial services respondents reporting quantifiable positive ROI from generative AI.

• Agentic AI is emerging as the next frontier, but governance will determine who scales it safely. While 30% of financial services firms have agentic AI in production, those using it report strong outcomes in analytics, forecasting, customer interactions and process automation — making controls around permissions, auditability, oversight and data access essential.

• AI is delivering positive workforce change — not just automation. Leading financial services companies lead other industries in positive workforce impact, with 78% of respondents saying AI-powered automation has had a net positive job impact.

• Proprietary data is becoming the foundation of competitive AI advantage. Ninety-two percent of financial services firms are training, tuning or augmenting large language models with proprietary data, underscoring the importance of trusted, governed enterprise data in creating differentiated AI outcomes.

For financial services leaders, the AI conversation has changed quickly. The question is no longer whether artificial intelligence will fundamentally reshape the industry. It already is. The more important question now is how financial services companies can turn AI into measurable business value while maintaining the trust, governance, security and control the industry requires.

New Snowflake research shows that financial services firms are entering a more mature stage of AI adoption. While individual firms vary on their adoption maturity, they are quantifying ROI, using proprietary data, applying AI to high-value enterprise workflows and beginning to capture early benefits from agentic AI.

The result is a compelling picture of an industry that has rapidly moved from AI curiosity to AI accountability.

AI is already creating a positive workforce impact

One of the most striking findings in the research is that financial services lead all compared industries in positive workforce impact from AI-powered automation. The compared industries include advertising and media; healthcare and life sciences; manufacturing; retail and CPG; and tech and telco. 

Seventy-eight percent of financial services respondents say AI-powered automation has had a net positive job impact. That is higher than tech and telco at 75%, healthcare and life sciences at 69%, advertising and media at 68%, manufacturing at 68.1%, and retail and CPG at 61%.
 

This finding challenges the common assumption that AI adoption is primarily about job elimination. In financial services, the story appears more nuanced. AI is changing work, but respondents are more likely to see that change as positive.

That makes sense in an industry where employees often navigate complex, repetitive, data-intensive and document-heavy workflows. AI can help them reduce manual effort across tasks such as summarizing documents, preparing data, monitoring compliance and automating help desk tickets.

For executives, this reframes the workforce conversation. The real opportunity is not simply replacing people with automation. It is about giving employees tools so they can work faster, make better decisions and focus on higher-value activities.

Financial services are measuring tangible ROI 

Companies across the industry are also among the most likely to connect generative AI adoption to measurable value. Sixty-eight percent of financial services respondents say they know gen AI ROI is positive because they have quantified it. Only tech and telco are slightly higher at 70%. Financial services is ahead of advertising and media at 64%, retail and CPG at 59%, healthcare and life sciences at 57%, and manufacturing at 56%.

This is an important signal. Financial services firms appear to be applying the same performance discipline to AI that they bring to other areas of the business.

That discipline matters. Quantified business impacts help leaders decide which use cases to scale, which pilots to stop, and where to allocate investment. It also helps prevent AI experimentation from becoming fragmented or disconnected from business strategy.

For financial services leaders, the message is clear: AI strategies should be built around measurable business outcomes, not technology novelty.

Proprietary data is becoming the foundation of AI advantage

In financial services, data has always been a strategic asset. AI raises the stakes by making proprietary data central to an enterprise’s AI differentiation. Ninety-two percent of financial services respondents say their organizations are training, tuning or augmenting large language models with proprietary data. 

This places financial services firmly among the advanced AI adoption tier, alongside tech and telco at 96%, healthcare and life sciences at 93%, advertising and media at 93%, retail and CPG at 91%, and manufacturing at 88%.

This is a critical shift. Generic AI tools can generate generic outputs. But interoperable AI systems grounded in proprietary enterprise data can produce more relevant, contextual and differentiated results.

For financial services firms, that could mean AI systems that support:

• Customer-specific service recommendations

• Risk and fraud analysis

• Internal policy and compliance guidance

• Market and scenario analysis

• Personalized financial insights

• Product and portfolio intelligence

• Enterprise knowledge management

But proprietary data use also introduces greater responsibility. The more AI systems rely on sensitive enterprise and customer data, the more important it becomes to govern access, monitor quality, protect privacy and maintain transparency. 

In financial services, AI advantage and AI governance have to advance together.

Agentic is the next competitive frontier

Gen AI has forever changed how employees interact with information. Agentic AI changes what AI can do on their behalf.

Agentic AI systems can reason through tasks, use tools, coordinate steps and complete workflows within defined boundaries. For financial services, this opens the door to new forms of automation across risk and regulations, operations, compliance, fraud investigation, forecasting and strategic planning.

Interestingly, financial services is not the top industry for agentic AI already in production. Thirty percent of financial services respondents say they are using agentic AI in production today, compared with 42% in advertising and media, 33% in healthcare and life sciences, 32% in tech and telco, 32% in manufacturing, and 28% in retail and CPG.

But among organizations implementing the technology, financial services report some of the strongest outcomes:

• 94% say agentic AI has provided better data analytics and strategic recommendations

• 92% say it has enhanced forecasting and scenario planning with advanced modeling

• 91% say it has improved customer-facing interactions via agents

• 83% say it has replaced or reduced human involvement in repetitive, rule-based processes

• 71% say it has replaced or reduced human involvement in complex, sophisticated processes

These outcomes are especially relevant for financial services because so much of the industry’s work is data-intensive, process-heavy and decision-oriented. The next level of agentic workflows span a wide range of use cases, from helping teams manage multi-step customer workflows and investigating fraud to streamlining onboarding and supporting scenario planning.
 

The key will be data governance. As AI systems move from generating answers to taking actions, financial services firms will need strong controls around permissions, auditability, human oversight and data access. For instance, sensitive data needs to be classified and lineage captured automatically; access policies must be enforced at the platform level not at the individual software or application level. Additionally, firms that implement an agentic control plane or an agentic mission control center will have an advantage. This layer is critical as it helps mitigate operational risk by coordinating data, models and applications to ensure agents work toward shared objectives within clearly defined guardrails. 

Data readiness and context will determine who scales AI successfully

The most mature AI strategies will depend on the quality, accessibility and governance of enterprise data, combined with the business context or semantics of the data. AI outputs are only as strong as the data and the business meaning behind that data combined.

An agent that understands the difference between a Know Your Customer verification workflow and a trade settlement exception is one your teams will actually adopt, driving tangible value across your organization.

Financial services leaders appear to especially recognize the data access issue. Ninety-six percent of respondents agree their organizations are actively investing in solutions to unify or consolidate their data estate. In addition, 89% agree that data engineering is key to AI project success.

But challenges remain:

• 62.8% agree their organization has a data engineering skill gap

• 57% agree AI initiatives are slowed by fragmented data systems and silos

• 51% agree their organization lacks visibility into the entirety of its data estate

This is the strategic bottleneck and the strategic opportunity. Financial services firms may have clear AI ambitions, but fragmented data limits how far AI can scale. If data is siloed, inconsistent or difficult to govern, AI systems will struggle to deliver trusted outputs and reliable, compliant actions.

For agentic AI, this becomes even more important. Systems that take action on behalf of users need access to accurate, governed contextual data. Without that foundation, firms may struggle to move from isolated use cases to enterprise transformation.

The next phase of financial services AI leadership

The financial services industry has unequivocally entered a new phase of AI adoption. The data shows an industry that is increasingly mature, disciplined and outcome-oriented. But the next phase of agentic workflows will require a more comprehensive approach. 

It will require trusted data foundations, strong governance and semantics, enterprise-grade tools, workforce enablement and consistent measurement.

For executive decision-makers, the path forward is clear:

• Prioritize AI use cases with measurable business value

• Ground AI systems in governed, contextual proprietary data

• Invest in interoperable data platforms that bring leading AI models to the data

• Build governance models and implement an agentic control plane for agentic AI before scaling it broadly

• Treat AI transformation as both a technology strategy and an operating model shift

The firms that lead will not be those that deploy AI everywhere at once. They will be the ones that connect trusted data, governed execution and measurable outcomes. In financial services, the future of AI will be quantifiable, highly governed and increasingly agentic.

To read the complete research findings, download The ROI of Gen AI and Agents 2026 now.