The Next Phase of Enterprise AI Is About Decisions, Not Experiments

At this year’s Milken Institute Global Conference, the discussions around enterprise AI had shifted. The question among executives was no longer if AI would pay off, but how. Many talked less about proving ROI and more about positioning their companies to capture it.

That tension—between AI’s capabilities and organizations’ capacity to keep up—was evident everywhere, from panel discussions to a private dinner co-hosted by Aily Labs founder and CEO Bianca Anghelina and The Information’s Cory Weinberg.

The offshoring analogy that won’t quit

Private equity firms announced a wave of joint ventures with frontier AI labs just as Milken opened—a timely reframe for an industry eager to justify its hefty valuations, even as harder questions about the workforce went largely unresolved. The historical parallel executives kept reaching for: offshoring.

Like offshoring before it, AI is reshaping labor across healthcare, finance, insurance and customer service, with near-term returns more incremental than transformational. KKR’s Pete Stavros put it plainly: earnings across his portfolio were up around 5%, not the 50% headlines imply. General Atlantic’s Martín Escobari was more bullish, arguing the technology is ready to automate most white-collar work, and what’s lagging is “prioritization and diffusion.”

Designing the AI decision layer

At the dinner, executives who described winning with AI shared one instinct: treat it as a decision-making layer across the business, not a collection of tools. One example came from gem certification: following Blackstone’s acquisition of IGI, AI-powered image detection systems are being used to identify lab-grown diamonds. At roughly 95% accuracy, they’re outperforming human reviewers in certain classification tasks.

Healthcare executives described a growing comfort using LLMs to interpret bloodwork, surface possible diagnoses and support preventive care conversations.

That thinking sits at the core of what Aily Labs is building: the first AI-native decision intelligence platform for global enterprises, designed to move Fortune 500 companies from fragmented data to measurable P&L impact in less than two weeks.

Aily’s platform orchestrates autonomous AI agents that do more than just surface insights—they execute decisions, with impact that runs to the top and bottom line across finance, supply chain, R&D and commercial operations. The goal, as Anghelina framed it, is to give every level of the organization the same real-time intelligence and the ability to act on it.

The distinction between AI that surfaces an insight and AI that takes action came up repeatedly at the dinner. Anghelina has argued that this is where most enterprise deployments stall—a system flags an inventory problem, and then a human has to pick it up from there. Aily’s approach goes further: its inventory agent traces overstock or understock risk to the root cause and acts autonomously to mitigate it before it hits the P&L. In one Fortune 500 deployment, that translated to $685 million in unlocked inventory value.

“In an enterprise context, real agents don’t stop at a recommendation. They take action,” said Anghelina.

The adoption gap

One observation to emerge from the dinner was the distance between how C-suite leaders are using AI and how their organizations are. Several executives described using AI as an always-on collaborator—synthesizing information across the P&L, stress-testing decisions and accelerating work that used to take days.

Anghelina was direct about what that gap reveals. “The question isn’t whether AI belongs in the boardroom anymore,” she said. “It’s why it’s still stuck there.” The goal, she argued, isn’t just limited to better insights for leadership, it’s a single source of real-time intelligence that every employee, from the CEO to the front line, is operating from.

But the picture looks very different a few floors down, where large portions of the workforce remain early in the adoption curve. That gap is precisely what decision intelligence platforms like Aily are designed to close, not through mandates, but by embedding AI into the daily routines where employees already work, making it easy enough to use that they find the value themselves.

The harder question

One question went unresolved: what happens to people whose jobs change, or disappear. Executives noted a telling countertrend: consumers gravitating toward more analog, human experiences even as AI embeds itself deeper into daily life.

The most compelling argument at the table wasn’t that AI would replace human judgment altogether, but that it would increasingly shape how judgment gets exercised inside organizations. That’s the part that can’t be offshored.