A recent Box survey of 1,300+ IT leaders found that 94% of organizations are using AI — yet the research reveals a clear divide. While most companies are experimenting with AI, the true leaders are early adopters, using AI to transform their operations and realize real results.
What sets these pioneers apart? They’ve integrated AI deeply into core operations and strategically prioritized AI investments.
The difference isn't who is using AI — it's how they're using it.
Read on for key insights from our State of AI in the Enterprise report.
Key takeaways of how leading-edge AI adopters are using AI more effectively:
- Leading-edge AI adopters are seeing 37% productivity improvements on average
- They’re using AI for more advanced use cases like process automation, metadata extraction, and cybersecurity
- They’re more likely to use AI agents for fully autonomous operations
- They think a lot about which model or agent is best for each use case rather than just using what’s easily available
They’re spending ~25% of their IT budget on AI
Measuring the ROI of strategic AI adoption
The immediate benefits of AI adoption often include time savings and increased productivity from automating routine administrative tasks like data entry, document processing, scheduling, and reporting. Cost reductions (and error reductions) naturally follow a lot of these efficiencies. The impact of AI also extends to customer satisfaction from factors like enhanced personalization, faster resolution times, and consistent service quality across channels.
But for the AI early adopters — those who aren’t just experimenting with AI but embedding it strategically across multiple business processes — the benefits of AI extend well beyond incremental savings. One of the most striking findings from this research: Companies on the leading edge of AI adoption are experiencing average productivity improvements of 37%. This contrasts sharply with organizations still in the early stages of AI usage or those applying it only superficially, who see less of a difference.
Unlocking the full potential of AI with advanced use cases
As organizations mature in their AI journeys, their use cases evolve from simple applications to more complex, value-generating ones. Writing communications, document analysis, and AI chat and research are still the most common use cases of generative AI, but that’s changing. The companies achieving higher AI ROI now employ advanced capabilities.
The leading-edge adopters are harnessing AI to accelerate information retrieval and synthesis, enabling knowledge workers to access critical insights in seconds rather than hours. They’re streamlining approval workflows and decision-making processes to eliminate costly bottlenecks. They’re leaning into AI for a level of error prevention that reduces expensive rework by ensuring tasks are done right the first time.
They’re also using metadata extraction to transform their unstructured data — marketing campaign videos, sales contracts, invoices, internal process information, supply chain data, and other proprietary content — into a gold mine of information. For example, they’re building comprehensive knowledge-mining systems that extract concepts, relationships, and insights from vast collections of data, creating searchable knowledge bases that preserve institutional memory and speed up information retrieval.
They’re also using AI-powered cybersecurity solutions that adapt through continuous learning to provide highly sophisticated data security. In regulated industries, they’re using AI to automate compliance by identifying sensitive information like personally identifiable information (PII), protected health information (PHI), and financial data, reducing the burden of manual reviews.
Agentic AI and model choice are key in the modern workforce
Many of these advanced use cases are enabled by AI agents, which can be trained on a company’s unstructured data for specific context. The great majority (87%) of organizations are already using AI agents — and 41% of companies are using them for fully autonomous operations.
Early AI adopters, in particular, leverage fully autonomous AI agents to transform their businesses. One technology startup CEO shared that their company plans to integrate advanced AI agents into both critical internal processes and their product offerings within the next one to two years.
Leading organizations are gaining a competitive edge by deploying multiple AI models and agents tailored to diverse needs. Rather than relying on a single solution, these companies strategically build AI portfolios combining foundation models for broad capabilities, specialized models with deep domain expertise, and custom models designed for unique challenges. This multi-model approach not only enhances flexibility but also drives significantly higher productivity gains.
Notably, top-performing companies invest time and resources in carefully selecting the best model for each use case — 83% of those with over 50% AI productivity improvements consider model choice very important. By thoughtfully integrating various AI technologies, businesses can solve complex problems more effectively and unlock transformative results that set them apart in a competitive market.
This move reflects a broader trend among high-ROI companies, which are increasingly adopting more sophisticated AI use cases to drive significant productivity gains beyond basic applications like writing or chat. As Box CEO Aaron Levie describes, “The more we can create agents that understand your domain, a set of instructions, and your corporate knowledge in a safe way that inherently knows your permissions and access, that’s the context that makes AI systems extremely powerful, and it’s why we’re going to see a revolution in what you can do with your unstructured data.”
The acceleration of AI is just beginning
Companies with the highest AI ROI have already invested heavily, dedicating about 25% of their IT budgets to AI. Nearly half of these high performers plan to boost their AI spending by more than 25% next year. The pace of AI transformation is expected to quicken, with most organizations anticipating full AI integration within two years.
The difference between average AI users and high-performing adopters lies not just in how much they invest, but in how thoughtfully and broadly they apply AI. Early adopters move beyond simple automation to embrace complex, strategic use cases that drive productivity, reduce costs, enhance customer satisfaction, and fuel revenue growth. Their success stories demonstrate that AI is no longer optional — it’s a critical driver of competitive advantage in today’s fast-evolving business landscape.
Companies looking to thrive in the AI-first era of business can learn from these leaders, investing wisely and deploying AI in ways that unlock its full transformative potential.
Read the full report for more examples of AI early adopters and their winning approach to AI.
