Aaron Levie

Enterprise AI’s near-term bottleneck is organizational integration, not model capability. Aaron Levie says coding agents thrive because engineers are technical, autonomous, able to debug failures, and working on verifiable outputs; ordinary knowledge work instead spans less-technical users, fragmented data, legacy systems, and undocumented relationships. Diffusion from startups into large enterprises will therefore take “a number of years.”

Top-down AI mandates are producing misleading failure statistics and measurable theater rather than operational change. Martin Casado says the claim that 95% of large-company AI efforts fail is “clearly silly” because employees are already using ChatGPT effectively; he distinguishes that from centralized, consultant-led programs that lack operational alignment. Token-counting incentives make the distortion worse—Aaron Levie says he and his coworkers assign agents useless tasks because “you get whatever you measure.”

Integration, permissions, and change management remain the durable enterprise workload—and potentially a decades-long services market. Steven Sinofsky’s hard stop is that every company with 1,000-plus employees or more than 10 years of history contains “a massive [amount of] stuff that’s sitting there waiting to be integrated,” while “AI actually doesn’t help to integrate anything.” Levie argues this makes work by Accenture, Deloitte, and other systems integrators entirely logical: people must implement the agents that may later automate work.

The central architectural shift is to treat an agent as a worker with identity, onboarding, and bounded authority—not merely software embedded in another product. Casado argues that companies already spent 40 years designing interfaces and processes for messy, nondeterministic humans, so an enterprise can “hire the agent,” give it email and application access, and reuse those controls. The unresolved problem is context: agents can operate at enormous parallel scale but do not naturally know which Sally or Bob to ask when the documented system fails.

SaaS may gain machine seats, but the API-versus-browser path remains contested. Levie sees Salesforce’s “full headless” move as a bellwether and says machine use could reach 100x or 1,000x human activity; Sinofsky calls an agent “another seat” because sharing human credentials would be indefensible. Casado and Sinofsky favor an API/MCP/CLI-first approach with browser use when those interfaces fail, while Casado notes that agents may need ordinary Safari when headless browsers are blocked.

Agentic volume creates two separate risks: familiar infrastructure scaling and less-understood operational entropy. Sinofsky asks what happens when 10,000 employee agents each hit a SaaS system 500 times more often; Casado says caching and standard distributed-systems techniques can address that load. His deeper concern is that AI-generated code “gets worse over time,” potentially creating as many problems as solutions, and companies do not yet know how to govern long-running agents whose output continuously changes shared systems.

The speakers expect AI to expand software, infrastructure, and skilled employment before it eliminates them. Box saw AI produce roughly 80%–90% of one feature, yet security review still constrained release; Levie therefore estimates perhaps 2x–3x engineering productivity, not 5x–10x. More code creates more systems to secure, upgrade, and repair, while John Deere, Caterpillar, Eli Lilly, and thousands of other companies can employ engineers using Claude Code, Codex, and Cursor: “We’re just getting started with the jobs on this front.”

Levie’s central labor bet is that AI will produce more engineers and lawyers over the next five years, not fewer. Tech represents perhaps 8-15% of GDP; giving the other 85% Silicon Valley-grade engineering shifts developers toward John Deere, Caterpillar, and Eli Lilly. AI also floods lawyers with draft contracts and memos while courts, patents, and professional approval remain constrained: “We haven’t removed humans from the loop. We’ve just changed where they enter the loop.”

Enterprise adoption’s bottleneck is organizational redesign, creating perhaps 500,000 to 1 million “agent operator” jobs. These technically fluent operators will understand MCPs, CLIs, skills, and AGENTS.md while rebuilding regulated workflows around agents. The resulting software workload is less another phone app than “an unlimited amount of software” connecting fragmented data and automating background processes.

SaaS is not uniformly cooked: agents strip value from button-heavy interfaces but amplify systems rich in APIs, proprietary business logic, governed data, and human review. An ERP is more than its database because supply-chain and accounting logic remain valuable; Box similarly says agents could drive API activity up by potentially 100x or 1,000x. Levie’s mandate: become “the best place where agents want to work with that data.”

Token spend escapes the IT budget and becomes regular operating expense, potentially doubling global technology spend rather than increasing it tenfold. Enterprises could trade a marketing campaign against automation, allocate premium models to the highest-value 5-10% of users, or make teams pitch for compute like contestants on Shark Tank. Unlike startups, large enterprises cannot “token max” indiscriminately: they have annual budgets and, for public companies, EPS commitments.

AI creates a structurally larger cyber market because agents both generate more vulnerable code and let attackers scan faster. As AI moves from writing most code toward 90% or 95%, every additional feature becomes another opportunity to open the wrong port or introduce a flaw. Defensive agents can review that output, but Levie’s formulation captures the circular opportunity: “Agents are the solution to the problem that agents have caused.”

Public software is being treated as an indiscriminately bucketed sector, but Levie expects separation over the next one to two years. Three-times free cash flow can be “aggressively low” for companies that respond effectively, although parts of software were previously valued beyond plausible terminal outcomes; winners will possess deep workflow ownership, valuable data, agent-ready APIs, and monetizable automation. He views Atlassian as potentially oversold because more engineering should expand demand for engineering infrastructure, even as its products must evolve.

Levie would still “load up” on frontier-model rounds despite the host’s test of an $850 billion valuation needing roughly a 3x path to $2.1 trillion. His cloud analogy is AWS at $500 million of revenue in 2010 versus a couple-hundred-billion-dollar ecosystem 15 years later: markets that work can become vastly larger than expected. OpenAI and Anthropic can both win in a multi-model world, while cross-lab infrastructure such as agent evaluations may spawn “a dozen, two dozen, five dozen” new categories.

Enterprise AI demand is running far ahead of cloud-era enthusiasm, but production remains in “very early innings.” Cloud required reluctant companies to surrender physical infrastructure and trust unfamiliar vendors; AI instead has executives proposing “almost as many use cases as possible,” sometimes more than practical. Levie sees a relatively narrow window in which large technology companies, enterprise-software vendors, and new agent startups can capture that demand.

AI changes IT from a software-enablement function into an operator of digital labor. Business units will ask IT not merely to deploy CRM or HR systems, but to provision agents that run sales campaigns, review contracts and invoices, or execute onboarding. Borrowing Jensen at NVIDIA’s framing, “the IT department becomes the HR department of AI,” requiring much deeper business knowledge and strategic authority.

Box’s RAG advantage came from architecture it built before ChatGPT, partly through luck. Box Hubs lets users curate authoritative documents without copying or changing permissions, reducing the risk that draft-heavy corporate repositories contaminate retrieval when users query a bounded hub. Because enterprises lack the public web’s PageRank-like authority signals, Levie argues this human curation makes Box’s RAG “at least a hundred times better” than indiscriminate search across all company data.

The strategic destination is a new “system of intelligence” that combines structured control with probabilistic judgment. Box’s current agents include model access, tools, skills, instructions, and enterprise data, though Levie concedes many would have been called assistants two years ago. The larger prize is agents that review content, make context-dependent decisions, route work, and coordinate with Salesforce, ServiceNow, Microsoft, or humans.

Being human-equivalent or 40% cheaper is insufficient to overcome enterprise adoption friction. A buyer still faces 17 competing projects, AI-council approval, testing that might take six months, privacy and security concerns, and workforce transition; consequential workflows may also demand “99.99999% reliability,” not 98%. Levie’s commercial threshold is an order-of-magnitude gain in cost, quality, or capability, especially for work that was never automated before.

Incumbents should retain their natural application domains, while startups win in cross-platform or previously unserved workflows. An “AI-first CRM” must assume Salesforce becomes AI-first too, just as Workday and ServiceNow will defend their domains; thin layers over an incumbent or OpenAI are therefore vulnerable. Independent compliance, agent red-teaming, cross-application workflows, and products requiring substantial non-AI interfaces remain credible startup territory.

AI may reopen SaaS pricing while improving productivity before the aggregate statistics clearly register it. Levie expects outcome pricing, compute consumption, and subscriptions all to coexist after two decades dominated by per-seat SaaS; Box itself is still testing. Internally, coding-tool gains range from 5–10% for some engineers to possibly 50% for a new hire, but Levie would take the over on economy-wide productivity forecasts if the clock starts in a few years.