Business leaders and IT decision-makers who have been cautiously investing in artificial intelligence just received one of the strongest signals yet that the technology is paying off. A sweeping analysis of more than 17,000 corporate earnings calls and conference transcripts conducted by Morgan Stanley found that 40% of companies actively deploying AI reported measurable financial gains during the second quarter. The finding, released in a research note this week, marks a notable shift from the hand-wringing of even a few months ago, when many wondered whether generative AI would ever move beyond promising pilots to real-world returns.

The hard numbers behind the optimism

Morgan Stanley analysts combed through transcripts spanning industries from manufacturing to financial services, searching for concrete mentions of AI-driven revenue growth, cost savings, or productivity improvements that could be tied directly to the bottom line. The 40% figure represents companies that not only talked about AI but could point to a quantifiable impact—not just experimental rollouts or vague efficiency claims. While the bank did not disclose a full list of the firms, its researchers noted that the gains were spread across sectors, suggesting that AI’s reach is broadening beyond the tech giants that first championed it.

Crucially, the analysis focused on the second quarter, a period when many organizations were still scaling their AI efforts. That 40% of adopters already see a return implies that the technology is maturing faster than typical enterprise software, which often takes years to show value. Morgan Stanley’s team emphasized that the finding is based on publicly available statements, meaning the actual number could be higher once internal metrics that companies don’t discuss openly are considered.

What this means for Windows-focused businesses

For the millions of organizations that run on Windows and the Microsoft ecosystem, the Morgan Stanley data carries specific weight. Microsoft has embedded AI deeply into its products over the past 18 months, from the Copilot assistant in Windows 11 to the Copilot for Microsoft 365 suite that spans Word, Excel, Teams, and more. Many of the companies that reported Q2 gains likely rely on these tools to some degree, and Microsoft itself has been touting customer success stories. Though the Morgan Stanley study didn’t single out Microsoft customers, the overlap is inescapable: Windows-based enterprises are among the largest cohort of AI adopters globally.

For IT administrators, the finding validates the pilot programs many have been running. “Measurable returns” could mean reduced time spent on routine tasks, faster data analysis in Excel, or fewer help-desk tickets thanks to AI-powered self-service. For business owners, it translates into a clearer business case for expanding Copilot licenses or building custom AI solutions on Azure. The study suggests that the window for skepticism is narrowing—companies still sitting on the sidelines risk falling behind competitively.

How we got here: a timeline of AI investment and doubt

The current milestone didn’t arrive overnight. Corporate AI spending surged after OpenAI’s ChatGPT launched in late 2022, with Microsoft quickly integrating the technology into Bing, Windows, and Office. Throughout 2023, enterprises rushed to experiment, but by early 2024, a wave of “AI fatigue” set in. Surveys from Gartner and others showed that while adoption was high, measurable ROI remained elusive for most. Chief information officers complained about the cost of AI infrastructure and the challenge of proving value to the board.

Microsoft responded by aggressively packaging AI features into existing subscriptions. The Copilot for Microsoft 365 add-on, priced at $30 per user per month, required a leap of faith from customers already paying for E3 or E5 licenses. Meanwhile, Windows 11’s Copilot button appeared on new PCs, and the company began charging for advanced AI capabilities in its cloud. By mid-2024, enough time had passed for early adopters to accumulate usage data and begin quantifying results. Morgan Stanley’s analysis captures that inflection point.

Other tech giants show a similar pattern. Google, Amazon, and Salesforce have all pushed AI into their platforms, and each has cited customer wins in earnings calls. But Morgan Stanley’s systematic review of 17,000 transcripts adds a layer of rigor that isolated anecdotes lack. It also aligns with recent commentary from Microsoft CEO Satya Nadella, who has said that AI adoption is happening “faster than any technology cycle” he has witnessed.

What to do now: practical steps for Windows IT teams

If 40% of peers are already seeing gains, the practical question for Windows administrators is how to join that group—or accelerate their progress. The Morgan Stanley research doesn’t provide a playbook, but industry best practices and Microsoft’s own guidance point to several immediate steps:

  • Audit your existing AI usage. Start by inventorying where AI is already in play. Are employees using Copilot in Edge? Have teams experimented with Microsoft Copilot for Security? Understanding baseline adoption is critical before you can measure improvement.
  • Define clear metrics. Vague goals like “make people more productive” won’t cut it. Tie AI usage to specific key performance indicators: time saved per task, reduction in manual data entry errors, faster resolution of support tickets. The companies that reported Q2 gains likely had these metrics in place.
  • Run controlled pilots. If you haven’t already, deploy Copilot for Microsoft 365 to a subset of users and compare their output against a control group. Focus on roles where text generation, data analysis, or meeting recaps are major time sinks.
  • Leverage Microsoft’s admin tools. The Copilot Dashboard in the Microsoft 365 admin center and Windows Update for Business reports can show usage patterns and potential impact. These tools can help build the internal business case for wider rollout.
  • Re-evaluate your licensing. Microsoft frequently adjusts its AI packaging. For example, the company recently made some Copilot features available to smaller businesses and introduced consumption-based pricing for certain AI services. Make sure you’re not overpaying for unused capabilities.

Beyond the Microsoft ecosystem, the Morgan Stanley data suggests that AI is becoming table stakes in many industries. IT leaders should be having serious conversations with line-of-business executives about where AI fits into strategic plans for the rest of the year. The companies reporting returns are often those that integrated AI into core workflows rather than treating it as a side project.

A note of caution: not all gains are equal

While the 40% figure is encouraging, it also means 60% of AI adopters have yet to see measurable returns. Morgan Stanley’s research doesn’t break down why some succeed while others struggle, but common themes from related studies point to data quality, employee training, and process redesign as differentiators. AI tools like Windows Copilot are only as effective as the data they can access and the willingness of staff to change how they work.

Additionally, “measurable gains” could mean different things to different organizations. A small productivity bump that delights a mid-sized firm might barely register for a global enterprise. The Morgan Stanley analysis aggregates these disparate outcomes, so individual experiences will vary. The key takeaway isn’t that AI is a guaranteed win, but that a critical mass of companies is now extracting value—and the number is growing.

The road ahead: what to watch in the coming quarters

Morgan Stanley’s analysis is a snapshot, not a forecast, but it sets the stage for several developments to watch. First, the next round of earnings calls will reveal whether the 40% figure holds or climbs. Microsoft’s own fiscal Q1 report, expected in late October, will offer a direct look at how its AI products are performing, especially as Copilot moves beyond early adopters. Second, the upcoming Windows 11 24H2 update, which is expected to introduce more AI features and hardware optimizations, could accelerate adoption on the PC side. Finally, any regulatory moves around AI, particularly in the European Union, could affect how companies measure and report returns.

For now, the message for Windows users and IT pros is clear: AI is no longer just a laboratory experiment. The tools are in your hands, and a growing number of your peers are already using them to measurable effect. The question is whether you can turn that opportunity into your own Q3 success story.