this year’s morgan stanley The lineup of technology, media, and communications conferences overlapped, with Dario Amodei and Sam Altman (both central figures in the Pentagon drama), Jensen Huang, Satya Nadella, and dozens of enterprise software CEOs trying to convince investors that they can survive AI calculations.
I spent a day on the ground reporting, and then spoke with David Chen, head of global technology investment banking at Morgan Stanley, to unpack what I’d heard.
Chen said the biggest change from last year was the nature of the questions. In 2025, companies were talking about leveraging artificial intelligence to reduce expenses, such as co-pilots and automation tools to cut operating costs by several percentage points. Chen said it’s now a gamble.
“I think investors really didn’t want to hear about how people were being made more efficient by AI,” he told us. “They really wanted to ask: Are you a beneficiary or will AI threaten your entire business?”
And to their credit, companies tried to answer that question more directly this year. Enterprise software companies in particular have seen $1 trillion in market capitalization evaporate in just one week this year.
Chen drew a line between companies that use software to do critical things, such as calculating payroll or sending out invoices, where a 2% error is a serious problem, and companies that essentially organize public data and put it behind a nice interface.
The former still has a moat, he argued. The latter is facing serious problems.
“AI doesn’t kill software,” Chen said. “I’m rearranging it.”
But he didn’t sugar coat it. He addressed companies on the wrong side of this line, calling it “wartime, not peacetime.” And he made an interesting observation about leadership. In this environment, boards are starting to prefer product-oriented CEOs over sales and marketing types. If you need to reinvent your company’s backend to be AI-native, you need someone who understands architecture, not just pipelines.
CNBC producer Jasmine Wu coined a phrase that perfectly captures this change. The idea came from a conversation: “We moved from SaaS (Software as a Service) to SaaaS (Software for Agents as a Service).” box CEO Aaron Levy said earlier this week. He said agents are now his new customer base and he can see the business growing 10 times his existing business. The implications are huge: the software that survives will no longer be the software that humans use. This is the software used by the agent.
Asked whether infrastructure spending will increase or decrease in 2027 when it comes to ramping up AI, Chen said: “Probably at the same level.”
If so, the capital investment cycle for AI, at least by hyperscalers, may be nearing its peak.
Chen predicts that the balance of winners and losers will be rebalanced in the enterprise software space over the next year, with cybersecurity standing out as a segment with the right competitive moat characteristics to be a clear beneficiary rather than a victim of AI.
He also warned of a wave of next-generation semiconductor and systems companies focused on solving the connectivity, computing and energy bottlenecks that are constraining the creation of AI.
Summary of the whole week: AI has gone from thinking this is going to be a big thing to realizing that this is already a big thing. So show that you accept it.
