Corporate America won't shut up about agentic AI, or AI in general
In fact, executives are saying the word “AI” more than they’re saying “earnings” on earnings calls.
At its annual developer conference Monday, Nvidia CEO Jensen Huang unveiled a jaw-dropping number that briefly sent the chipmaker’s shares soaring: at least $1 trillion in expected revenue from its next-generation AI chips through 2027.
The demand, Huang said, is largely driven by the rise of “AI natives” like OpenAI and Anthropic, which have pushed computing demand for Nvidia’s GPUs “off the charts.”
Indeed, boardrooms have been consumed by AI for a while now, with executives increasingly talking about the two-letter technology more than the results they’re there to discuss. According to Bloomberg data, S&P 500 executives said the word “AI” nearly 5,000 times on S&P 500 calls in the first quarter alone, outpacing the word “earnings” by more than 1,200 mentions.
But for C-suite folks looking to juice their stock price, it's not enough to be talking about chatbots or generative AI. The AI term you have to work into your conference call spiel is agentic AI. That is, systems that don't just answer your questions but actually do things for you, from booking a meeting to filing an expense report.
Pointing to OpenClaw — the viral open-source tool that lets anyone build and run AI agents — Huang called it “the new computer,” adding that “every company in the world today needs to have an OpenClaw strategy.” Nvidia, for its part, unveiled its own NemoClaw on Monday, a more secure, enterprise-ready version that allows companies to deploy agents safely.
And corporate America is catching on fast. While mentions of “AI agents” and “agentic AI” on S&P 500 earnings calls were virtually nonexistent until late 2024, now they’re in the hundreds, rising more than fivefold over the last five quarters to 245 mentions in Q1 2026, per Bloomberg data.
Reducing agents
Wall Street likes talk of "agents" because the implication is fewer employees. Just recently, we've seen a big round of layoffs at Block, 10% of jobs slashed at Atlassian, and reports of huge cuts at Meta, with AI-powered efficiency gains promised in each case.
Mastercard recently launched an agentic AI tool to provide small businesses with C-suite level solutions. Meanwhile, JPMorgan Chase’s chief analytics officer told CNBC last September that the bank’s end goal is one where “every process is powered by AI agents,” with internal demos already generating a full investment banking deck in 30 seconds. Major retailers have also jumped in, with Walmart, Target, and Home Depot having partnered with tech providers to deploy agentic AI tools across their operations, from pricing to inventory. PepsiCo, maker of Cheetos, Lays, and Mountain Dew, wants to be completely agentic first.
Even the companies most bruised by the agentic AI wave are trying to embrace it. After a massive sell-off in software stocks earlier this year on fears that AI agents could displace traditional Software-as-a-Service (SaaS) models, Nvidia said on Monday it’s teaming up with a slate of software firms — including Adobe, SAP, and Salesforce — to build and run AI agents using the chipmaker’s Agent Toolkit platform. If you can't beat 'em, join 'em. Or at least ask your AI agent what to do next.
