---
title: "Unlocking enterprise AI: opportunities and strategies"
customer: "Databricks"
customerWebsite: "https://www.databricks.com/"
datePublished: 2026-09-09
dateModified: 2026-09-09
sourceAssetType: "report"
topics: ["enterprise-ai", "generative-ai", "data-governance", "data-modernization", "ai-strategy"]
canonical: "https://theintelligencelibrary.com/databricks/unlocking-enterprise-ai-opportunities-and-strategies/"
---

# Unlocking enterprise AI: opportunities and strategies

Enterprise AI delivers value when companies pair GenAI with their own data, modernize infrastructure, and add strong governance.

## Key takeaways

- 85% of organisations are already using GenAI in at least one function, but only 1 in 5 say current investment is enough.
- The report argues that enterprise AI value comes from integrating GenAI with proprietary data on a modern data platform.
- Only 22% of organisations say their current architecture can support AI workloads without modification.
- Production readiness is still low: 37% of executives and 29% of practitioners say their GenAI applications are ready.

## What is Unlocking enterprise AI: opportunities and strategies, and how was it researched?
Unlocking enterprise AI: opportunities and strategies is an Economist Impact report commissioned by Databricks. It combines a global survey of 1,100 respondents with interviews from 28 C-suite executives to explain how enterprises can move from AI experimentation to scaled deployment.

The survey includes 715 technical executives and 385 data and AI technologists. It was fielded from July to August 2024 in 19 countries across North America, Europe and Asia-Pacific. The interviews span 11 industries, including financial services, healthcare and life sciences, retail and consumer goods, the public sector, manufacturing, transport, energy, technology, professional services, media and entertainment, and telecommunications. Most interviewees were from multinational companies with headquarters spanning ten countries across the Americas, Asia-Pacific and Europe.

The report opens against a backdrop of an estimated US$1trn in capital expenditure needed in coming years to build the data centres, chips, energy and infrastructure behind the AI build-out.

## What is the biggest factor in making enterprise AI actually deliver value?
The report says the biggest factor is pairing GenAI with an organisation’s own data and know-how on a modern enterprise data platform. Databricks president of global field operations Andy Kofoid says winning organisations will take a holistic approach that includes data management, security, governance, culture and domain-specific expertise.

The report also says the future lies in true data intelligence: AI systems that understand and operate on an organisation’s specific, often sensitive datasets. That approach matters because enterprise AI value comes from context, not just from access to a large model trained on web data.

## How are large enterprises moving from AI experimentation to practical use across the business?
Large enterprises are already moving quickly, but most are still early in the shift from pilots to broad deployment. The survey found that 85% of organisations are actively using GenAI in at least one business function, and that figure rises to 97% among companies with revenue above US$10bn.

Adoption is strongest in IT, which is the most avid user. Other heavily active functions include:
- marketing at 85%
- sales and customer service at 83%
- operations at 80%
- data scientists, 82% of whom report using AI for coding

The report says 7 in 10 executives see AI as crucial to long-term strategic goals, while only 18% say it is overhyped. By 2027, 99% of executives expect GenAI adoption across both internal and external use cases. In the near term, organisations are using AI mainly for productivity and efficiency, but they expect longer-term gains from revenue, business model innovation and market positioning.

## Why does modernizing the data platform matter more than choosing one specific AI model?
Modernising the data platform matters more because enterprise AI performance depends on access to reliable, connected, governed data. Model choice alone cannot solve fragmented systems, weak data connections or limited visibility into internal datasets.

The report finds that only 22% of organisations say their current architecture can support AI workloads without modifications. It also finds that 48% of data engineers spend most of their time resolving data source connections, which shows how much time is still lost to basic plumbing.

As AI models become more commoditised, the report says the best performers will mix and match open- and closed-source models with their own proprietary data and know-how. Nearly seven in ten organisations are already experimenting with or have fully deployed open-source GenAI, and 96% believe they will do so by 2027. At the same time, 45% of data scientists are using large language models without retrieval augmented generation, which reduces control and makes enterprise use harder to govern.

## What data management and governance steps does the report recommend before broad rollout?
The report recommends fixing the data foundation before scaling AI broadly. That means reducing fragmentation, improving data source connections, and making sure enterprise AI can work across existing systems without creating more congestion and opacity.

It also means putting strong security and governance in place before sensitive data is exposed to broader use. Andy Kofoid says organisations need to avoid moving too fast, but they also should not move too slowly, because overcaution can block innovation just as easily as poor controls can create risk.

The report adds that enterprise AI decisions and outputs must comply with existing and emerging rules around data privacy, security and consumer protection. It also says companies need a workforce strategy that properly calibrates humans versus machines and a clear ROI strategy, because revenue gains may take time to appear.

## How can companies give more employees access to AI while still keeping data secure and controlled?
Companies can democratise AI by widening access to data and AI tools while keeping strong controls around sensitive information. The report says organisations want more employees to be able to ask questions in plain language and get accurate answers grounded in enterprise data.

That approach only works if organisations protect sensitive data, certify accuracy and keep governance in place. The report says democratisation is a major driver for AI because it allows a wider range of employees to work with data and to build and refine AI models that reflect day-to-day business needs.

## What kinds of guardrails does the report say organizations need around AI use and data access?
The report says organisations need guardrails that cover security, governance, privacy, consumer protection and human-machine interaction. Those guardrails matter because AI systems can expose sensitive data, create compliance risk or produce outputs that are not ready for operational use.

Andy Kofoid says security and governance are essential in the new AI landscape, especially as organisations rush to implement AI solutions and regulators continue to evolve the rules. The report warns that companies can move too fast or too slowly, so the right answer is disciplined control rather than paralysis.

## What is the report’s overall strategy for enterprises that want to scale AI beyond pilot projects?
The report’s strategy is to treat enterprise AI as a full operating-model change, not just a model-selection exercise. Databricks and Economist Impact point to a combination of modern infrastructure, data governance, workforce planning, ROI discipline and careful design of human-machine interaction.

The report says the future lies in AI systems and agents made up of multiple components, using custom models optimised for specific tasks. That approach can improve accuracy, cost, performance and security, especially when enterprises connect those models to proprietary data.

For regulated industries, the report’s message is clear: strong governance and security are not optional. They are the foundation that lets companies innovate without losing control.

## Frequently asked questions

### What is this report based on?

It is based on a global survey of 1,100 respondents and interviews with 28 C-suite executives. The survey covered 715 technical executives and 385 data and AI technologists in 19 countries and ran from July to August 2024.

[Direct link](https://theintelligencelibrary.com/databricks/unlocking-enterprise-ai-opportunities-and-strategies/#what-is-this-report-based-on)

### What did the global survey reveal about enterprise AI adoption?

The survey shows GenAI is already mainstream: 85% of organisations are using it in at least one business function, 97% of companies with revenue above US$10bn are using it, and 99% of executives expect broader adoption by 2027.

[Direct link](https://theintelligencelibrary.com/databricks/unlocking-enterprise-ai-opportunities-and-strategies/#what-did-the-global-survey-reveal-about-enterprise-ai-adoption)

### What is the biggest factor in making enterprise AI actually deliver value?

The report says the biggest factor is integrating GenAI with an organisation’s own data on a modern data platform. That combination creates the data intelligence needed for more accurate, secure and domain-specific AI.

[Direct link](https://theintelligencelibrary.com/databricks/unlocking-enterprise-ai-opportunities-and-strategies/#what-is-the-biggest-factor-in-making-enterprise-ai-actually-deliver-value)

### Why does  modernizing the data platform matters more than choosing one specific AI model?

Modernising the data platform matters because fragmented data and weak connections limit AI more than model choice does. Only 22% of organisations say their current architecture can support AI workloads without modifications, and 48% of data engineers spend most of their time fixing data source connections.

[Direct link](https://theintelligencelibrary.com/databricks/unlocking-enterprise-ai-opportunities-and-strategies/#why-does-modernizing-the-data-platform-matters-more-than-choosing-one-specific-ai-model)

### What kinds of guardrails does the report say organizations need around AI use and data access?

The report says organisations need strong security, governance, privacy and consumer-protection controls, plus careful human-machine design. It also says sensitive data must be protected and AI use must comply with existing and emerging rules.

[Direct link](https://theintelligencelibrary.com/databricks/unlocking-enterprise-ai-opportunities-and-strategies/#what-kinds-of-guardrails-does-the-report-say-organizations-need-around-ai-use-and-data-access)

### How should regulated industries think about AI if governance and security are major concerns?

Regulated industries should treat AI as a governance-first programme, not a fast experiment. The report says they need robust security, careful data access controls and compliance with evolving regulation so innovation does not outrun oversight.

[Direct link](https://theintelligencelibrary.com/databricks/unlocking-enterprise-ai-opportunities-and-strategies/#how-should-regulated-industries-think-about-ai-if-governance-and-security-are-major-concerns)

### What is the overall strategy for enterprises that want to scale AI beyond pilot projects?

The overall strategy is to modernise infrastructure, mix the right models, connect them to proprietary data, and pair them with workforce, ROI and governance plans. The report argues that enterprises should scale AI deliberately beyond pilots rather than relying on a single model or one-off use case.

[Direct link](https://theintelligencelibrary.com/databricks/unlocking-enterprise-ai-opportunities-and-strategies/#what-is-the-overall-strategy-for-enterprises-that-want-to-scale-ai-beyond-pilot-projects)


---

[Download the report (PDF)](https://theintelligencelibrary.com/downloads/unlocking-enterprise-ai-opportunities-and-strategies.pdf?utm_source=theintelligencelibrary&utm_medium=referral&utm_content=unlocking-enterprise-ai-opportunities-and-strategies)

Published by [Databricks](https://www.databricks.com/?utm_source=theintelligencelibrary&utm_medium=referral&utm_content=unlocking-enterprise-ai-opportunities-and-strategies).