Box (BOX) Cisco AI Summit 2026 summary

AI adoption and workflow transformation

• AI is rapidly transforming engineering, with agents now capable of building entire software products, but adoption in other knowledge work areas lags due to complexity and context requirements.

• Coding benefits from verifiable outputs and standardized practices, while other fields like legal or sales face challenges with permissions, context, and unstructured data.

• Effective AI deployment in enterprises requires re-engineering workflows to provide agents with the right context, permissions, and data access.

• Organizations must adapt their processes to fit agent capabilities, not the other way around, to unlock AI's full potential.

• Early adopters willing to transform workflows can gain significant competitive advantages.

Enterprise use cases and value creation

• Enterprises are leveraging agents to process vast amounts of unstructured data, such as medical records, to improve decision-making and efficiency.

• Automating workflows with agents can unlock new revenue streams, cost savings, and reduce bottlenecks, but requires significant change management.

• Identifying underutilized data and applying agents to extract value is a key strategy for enterprise transformation.

• Success is measured by both general productivity gains and targeted reinvention of critical workflows, aiming for several high-impact wins within a few years.

• Streamlining processes, such as onboarding or contract management, can dramatically improve competitiveness and customer insights.

SaaS, agents, and commercial models

• SaaS is expected to grow as software becomes cheaper to build, increasing competition and driving prices down.

• Most organizations will not build their own core systems but will continue to rely on specialized SaaS for non-core functions.

• The rise of agents increases the value of systems that manage workflow, permissions, and data access, as these become critical for coordinating large numbers of non-deterministic agents.

• Pricing models are shifting toward consumption-based approaches for agent-driven tasks, with potential for hybrid models as adoption scales.

• Customers will likely prefer flexible, consumption-based pricing during experimentation, moving to fixed rates as usage stabilizes.