In a significant move to strengthen its position in the rapidly evolving AI landscape, Box has unveiled a comprehensive new AI platform designed to transform how enterprises interact with their content. The announcement, made on May 15, 2025, introduces powerful AI agents that will enable businesses to extract deeper insights from their data without moving content across multiple platforms.
Box’s Strategic AI Evolution
Box CEO Aaron Levie describes the current moment as a return to the company’s startup roots. “Box has never felt more like a startup than it does today other than the first 12 months of our existence,” Levie explains. The AI revolution has created a sense of urgency and innovation at the company that now employs 2,800 people and generated over $1 billion in revenue last year.
The company’s latest announcement centers on three key agent capabilities: a search agent that can find precise information across vast repositories of content; a deep research agent capable of analyzing multiple documents to synthesize insights; and a data extraction agent that can pull structured information from unstructured content like contracts and invoices.
“We built an AI platform that lets you take the breakthroughs from OpenAI or Google Gemini or Enthropic or Meta or XAI and bring that to your content so you don’t have to ship your data around to all these different platforms,” Levie explains.
Microsoft Partnership Expansion
Among the most notable aspects of Box’s announcement is its integration with Microsoft 365 Copilot, extending Box AI capabilities into Microsoft Teams, Word, PowerPoint, and Copilot Chat. This allows joint customers to leverage Box’s content management within Microsoft’s productivity ecosystem.
“Box has been a trusted Microsoft partner for years, and we’re pleased to take our collaboration to the next level with this new Box AI Agent for Microsoft 365 Copilot,” said Dan Stevenson, General Manager, Copilot Ecosystem at Microsoft. “By bringing together Box’s enterprise-grade Intelligent Content Management platform and Microsoft’s AI-powered productivity platform, we’re making it seamless for joint customers to use the power of Copilot with all their content stored in Box, unlocking insights and taking actions across their enterprise tools.”
Despite some competitive overlap, the partnership reflects Box’s long-standing approach to interoperability within the Microsoft environment. “We want to be the best place for bringing AI to content but we know that there’s lots of other things that you’re going to do with AI and we want to make sure we plug into those environments as well,” Levie notes.
The Reality of AI Agents in 2025
Levie takes a pragmatic view on the current state of AI agents, describing 2025 as “day one of agent adoption in the enterprise.” He suggests that organizations should focus on clear use cases where agents can deliver immediate ROI with high accuracy.
“We should think about 2025 for agents in the same way that we probably thought about 2023 and 2024 for just AI chat bots,” Levie states. “These are the first times that you’re going to see them in the enterprise and it’s going to start with a proof of concept for a small team.”
While adoption is still in early stages, he predicts exponential growth: “We will still have a hundred times more agents in the enterprise this year than last year, but in five years it’ll probably be something like 100,000 times the number of agents.”
Transforming Untapped Data Into Value
One of the most compelling aspects of Box’s AI strategy is its focus on extracting value from previously underutilized data. When showing these capabilities to customers across industries—from talent agencies to insurance companies—Box consistently discovers untapped opportunities.
“We’ll go to a talent agency and we’ll say ‘If you could understand everything that was inside of your contracts and your agreements with your talent, what would you do differently?'” Levie explains. “And instantly they’ll light up with new ideas about how they could generate more revenue.”
AI Creating More Work, Not Less
Contrary to concerns about AI eliminating jobs, Levie argues that AI is actually creating more work by enabling organizations to tackle previously infeasible projects. “The vast majority of use cases that customers will come up with… are use cases that humans were not doing before,” he notes.
“I’ve probably caused far more work for humans as a result of AI probably by an order of magnitude than I’ve actually saved in human time because of AI,” Levie admits. By accelerating project timelines and idea generation, AI enables companies to move faster and take on more initiatives than previously possible.
Levie predicts that “90% of the work that AI is doing in a couple years from now is work that humans never did before AI.” This perspective suggests that AI’s primary impact will be expanding capabilities rather than replacing existing roles.
As Box continues rolling out these new AI capabilities in the coming months, it represents a significant bet on how AI will reshape enterprise content management. For an industry veteran like Levie, who co-founded Box back in 2005 with fellow Mercer Island High School alum Dylan Smith, this latest evolution marks yet another transformation in the company’s journey from simple file sharing to intelligent content management.