This report was independently researched and produced by GILD ahead of the GILD AI and the Future of Internal Productivity event.

The research surfaces key patterns, decisions, and tradeoffs related to AI and the Future of Internal Productivity and is provided to attendees after the event to support continued reflection and application beyond the live discussion.

It is shared with attendees as a post-event resource to extend the value of the session and support continued application inside their organizations.

The GILD AI and the Future of Internal Productivity event was underwritten by Tecla.

GILD independently researched, authored, and produced this report and retained full editorial control, analysis, and conclusions.

How to Use This Report

This research report is published after each GILD event as an additional value-add for attendees and registrants. It is designed to help CEOs, founders, and COOs move faster inside their organizations on the topic discussed.

The core conversation stays private by design. This document stands on its own as a research-backed, operator-grade asset that leaders can confidently share internally.

This report:

• Is not a recap of the event
• Does not summarize or reference anything said in the room
• Does not include quotes, examples, or identifiable actions from attendees

All insights are synthesized from external research (late 2025–early 2026) and translated into practical guidance.

Executive Summary

AI has entered a new phase inside modern technology companies. Adoption is no longer optional, experimentation is no longer impressive, and leadership teams are increasingly judged on whether AI investments translate into real productivity gains.

"AI adoption is no longer the problem. The real gap is between experimentation and measurable productivity."

By late 2025, AI is embedded across most organizations, yet only a small percentage have achieved meaningful operational or financial impact. The gap between adoption and value is driven not by technology limitations, but by workflow design, operating models, and leadership decisions.

This report examines:

• The current state of AI-driven internal productivity
• The changes reshaping AI adoption and why they matter now
• How leaders are redesigning work for AI-native execution
• Where AI initiatives break and how to avoid common failure modes
• Practical frameworks, checklists, and experiments leaders can use immediately

The State of the World: AI and Internal Productivity (Late 2025)

AI usage inside technology companies has reached near-saturation. Most organizations now use AI in at least one internal function, and investment levels have grown at historic speed.

What has not kept pace is enterprise-wide productivity impact.

Key Observations

• AI adoption is widespread, but depth of use remains shallow
• Productivity gains are localized to teams or roles, not scaled across systems
• Leaders are under pressure to show ROI as AI spend accelerates

Enterprise GenAI Spend Growth

Despite this surge, value capture remains limited.

The gap between AI adoption and value capture defines the current enterprise landscape.

AI Adoption

Companies using AI in at least one function

Material EBIT Impact

Companies achieving EBIT impact from AI

What Is Changing, and Why It Matters

The AI productivity conversation has shifted from tools to systems.

From Experimentation to Expectation

AI initiatives are now expected to:

• Reduce cycle times
• Increase output per employee
• Improve decision quality
• Lower operational friction

Leaders are increasingly evaluated on whether AI changes how work gets done, not whether tools exist.

The Productivity Lag

Many organizations experience delayed gains due to unchanged workflows and operating models.

Organizations move through pilot excitement, reality dip, workflow redesign, and finally scaled gains.

The AI Productivity J-Curve

Y-axis: Productivity Impact
X-axis: Phase

The Rise of AI Agents

AI agents are increasingly used for multi-step internal tasks, particularly in engineering, IT, and operations.