At a company the size of PayPal, convincing tens of thousands of employees to adopt AI is no small feat. With teams spanning marketing, data science, risk management, and model control across 200+ markets, rolling out AI could have easily meant months of training slides, pilots, and partial adoption.

But PayPal took a different approach: It gave employees access to multiple AI solutions and let them pick what worked best for them. For Keith Swiderski, PayPal’s AI Fluency and Transformation Lead, accuracy was non-negotiable. To inform high-impact decisions, he needed accurate market intelligence and competitive analysis.

A short learning curve, a wide range of use cases

Model Governance Lead, Graham Woods, faced a different challenge: he needed an AI solution that could double-check PayPal’s in-house models without introducing new errors. Across the organization, the ask was simple: PayPal needed something reliable, efficient, and easy to learn that could handle everything from market research to data validation, with minimal rework.

PayPal’s approach was by design. For Prakhar Mehrotra, PayPal's Global Head of AI, "AI creates the most value when employees trust it enough to use it in decisions that matter. Our approach at PayPal has been to build that trust from the ground up, giving teams access to the best tools and trainings, letting results speak for themselves, and using what we learn to mature how each of us uses AI. The companies that win the next decade will be the ones that embedded AI deeply enough that it changed how they think and work. PayPal's goal isn't AI adoption for its own sake, it's building a workforce where AI fluency is expected, and where every team moves with the speed and precision that can change outcomes."

Adaptable across a wide range of workflows

PayPal employees found Perplexity to be consistent, accurate, and adaptable across a range of workflows. Today, Keith uses Perplexity Enterprise to swiftly distill competitive intelligence, channel data, and market trends into insights that move the business forward. When he built the internal case for new marketing opportunities, Perplexity delivered everything he needed including engagement data and industry benchmarks in a single conversation.

“Whether it's reporting on channel performance or the business sector as a whole, I know I can trust the outputs I'm getting,” he says. "Rather than validating hundreds of fragmented sources or leaning on research firms, I can pressure-test ideas in real time and walk into leadership meetings with recommendations I know are grounded in verifiable data. It’s fundamentally changed how we approach our work.”

For Graham, the tool has become a key “human-in-the-loop” partner in his validation process. He uploads transcripts, reviews Perplexity’s interpretation, and cross-checks it with PayPal’s internal model output. It makes it far easier to trace how a misclassification happened. Was it a logic flaw, a definition issue, or just messy data? Now, he can tell in minutes.

A clear trail for investigating further

Meanwhile, Enterprise Risk Monitor, Michael Abboud, uses Perplexity Enterprise to verify that people and businesses are exactly who they claim to be. His work includes cross-referencing ownership details, identifying mismatched entities, and flagging suspicious accounts. This can mean juggling multiple tabs and fragmented sources. Now, by using Perplexity, Michael works in one clean workflow, with citations that provide a clear trail when something doesn’t add up.

PayPal proved that large-scale AI adoption doesn’t have to move slowly. By giving employees access to a range of AI tools and the autonomy to integrate them into their own workflows, PayPal saw more confident decision-making across its teams.