AI Adoption: Takeaways From an AI Summit, and What Applies to You

AI Adoption

AI Adoption: Takeaways From an AI Summit, and What Applies to You

AI Adoption: Takeaways From an AI Summit, and What Applies to You

About this brief

What is the ECLC? The Executive Council for Leading Change (ECLC) is a global, peer-led community of more than 1,200 enterprise transformation leaders. Members meet in regular virtual roundtables of around 45 executives, held under Chatham House rules, to share what is working, what is not, and how they are handling change.


How Tigerhall fits. ECLC is convened by Tigerhall, the transformation delivery platform for a world of always-on change. Tigerhall turns a companyโ€™s strategy into personalized change journeys for each employee, delivered in the flow of work (Teams, Slack, Outlook, mobile) in 33 languages, with real-time data on where the change is landing and where it is stalling. Leaders own the strategy and the plan; Tigerhall runs the delivery, so a team of two or three can deliver what used to take fifteen to twenty.


What is an executive brief? An ECLC executive brief is a short, anonymised write-up of a single roundtable. It captures what the room discussed, where leaders agreed and differed, and what to do next. Each opens with a summary, then themed sections, then eight key takeaways.


Executive Summary for AI Adoption

A member-led roundtable took the lessons of a one-day, AI-only summit and tested them against the experience of about 30 transformation leaders. Using a live anonymous poll, the group put its own challenges in order: ownership and proving value came out on top, and most teams are already in rollout.


The discussion kept returning to one point: AI does not remove accountability, it concentrates it, so every process, tool and output needs a named human owner, a business metric and a route into governance.


What the Summit Surfaced

The summit brought together leaders from a wide spread of sectors, including financial services, professional services, energy, consumer goods and data infrastructure. Six themes ran through the day, and each was open to challenge in the room.

  • Where to start: Business problem first, or process first? Maturity differs widely.

  • Getting people to use it: Adoption advice varied, but it centred on habits, not demos.

  • Who owns it: Humans remain accountable for anything an AI process produces.

  • Proving the value is real: The most debated topic of the day.

  • What happens to the work: Where does freed-up time go, and what happens to roles?

  • Will it work here: AI amplifies the culture and processes already in place, including the broken ones.

The roomโ€™s own poll echoed this. Roughly 44 percent worked in change or workforce transformation, ownership was the top challenge at about 29 percent with proving value close behind, and most teams were already rolling out AI.


One Owner per Process

The summitโ€™s advice was blunt: give each process one owner, not a committee and not a central team. Members agreed with the principle but pushed on how it works in practice.

  • Decision rights are the real question. Owning a process is not the same as having the authority, or the appetite, to redesign it.

  • Executive sponsorship has to sit alongside it. Without alignment on who decides, ownership alone does not move adoption.

  • Process experts are not always change drivers. In organizations with long-tenured teams, the owner may be a subject-matter expert who needs a challenger.

  • Product-minded leaders help. One member put a product-experienced executive sponsor over cross-functional work and paired them with subject-matter experts.

  • Use a business-line sponsor with P&L accountability. Subject-matter experts connect process improvement and digitalization behind them.

  • Include a pressure tester for the human side. Someone should ask where humans re-enter the loop, and someone who will not simply fall in line.

Incentives came up repeatedly. How people are measured and recognized shapes who ends up owning what, so it needs settling early.


Proving AI Value

A facilitator from the summitโ€™s ROI discussion described different kinds of return an AI effort can target, each with its own timing and purpose: protecting the business, setting it apart, or improving efficiency or quality.

  • Start from the business metric. Revenue, margin or acquisition cost, with a clear hypothesis about what is blocking it.

  • Lean toward growth over cost-cutting. Growth stories are easier to align people behind, and they tie neatly to incentives.

  • Go narrow. One team found the hardest part of the sales process before deploying anything, rather than rebuilding the whole process.

  • Treat it as ordinary value modelling. One member building an AI value portfolio said the discipline is no different from any strategic initiative; companies that were already good at value stories are doing it well.

  • Frame it around the customer. Teams adopt faster when the internal benefit visibly reaches their customers.

  • Expect the how to move. Better models change timing and likelihood, but not the value of the outcome you are after.

One member added that AI differs in a meaningful way: work is moving to non-humans, which changes how you review, hold accountable and give visibility to it, and tasks drift in organic ways rather than on a planned transformation schedule.


Accountability and Over-Reliance

The room spent real time on the human side of working with AI.

  • Over-engagement is a risk. Complimentary, always-available tools can pull users into loops of rechecking with little output.

  • Models are getting less sycophantic, but not enough to stop reviewing their work.

  • Low-effort volume lands on colleagues. A long, unreviewed document pushes the review burden onto someone else, and the blame lands on the sender, not the tool.

  • Say it explicitly. Make clear that people remain accountable for what they send.

  • Write a discernment policy. One organization did this to cut low-quality output.

  • Protect independent thinking, especially for early-career staff who may lean on the technology before they have built their own judgement.

Working with AI is personal. Guardrails and guidance work better than prescribed ways of working.


Driving Adoption

The summitโ€™s adoption lesson was to begin with one task that employees already do, then build habits from there, instead of demonstrating a tool.

  • Involve the people in the process. Anyone who touches the process today should help decide what is automated.

  • Capture the impact first. Know how the work changes for each person before it is automated.

  • Plan for the governance load. Checking outputs and applying governance is ongoing work, and people who were not part of the automation are less willing to do it.

Build, Buy and Governing What People Build

One member raised the question of how to move a tool someone built for themselves into something the enterprise can rely on, and how to catch duplicate builds of the same use case.

  • Widen the technology review board to cover internal AI tools, applying the same security, legal and support scrutiny as external vendors.

  • Put the transformation or innovation lead on the AI governance body, so it balances rigor with energy.

  • Heavy governance can backfire. A member who ran a hackathon saw the review process become a disincentive; the reputation of five meetings a week spread fast.

  • Anchor on business risk, not prohibition. If a tool built on one personโ€™s account becomes critical to a process, losing that account is a business continuity risk.

  • Consolidate overlapping efforts. Where many people pursue the same goal, such as knowledge management, turn it into an enterprise project with resourcing and visibility.

  • Use crowdsourced use-case tracking so teams can see and vote on what should be built.


Jobs, Headcount and the Talent Pipeline

  • Growth can add roles. One leader uses AI to get past the point where a team needs to add headcount, expecting growth to eventually create more roles, not fewer.

  • The entry-level pipeline is thinning. Work moving to AI is, in part, work for employees who may never be hired, which raises later questions about mid-level and subject-matter expertise.

  • Succession planning matters. Building the pipeline from entry level to executive remains important.

  • Careers may look different: more micro-organizations, portfolio careers and a rethink of education.


Key Takeaways

  1. Name one human owner per process. Pair them with an executive sponsor and subject-matter experts.

  2. Clarify decision rights early. Ownership without authority stalls redesign.

  3. Start value work from a business metric. Tie each AI effort to revenue, margin or cost, and be explicit about the type of return.

  4. Favor growth framing. It is easier to align people behind and links better to incentives.

  5. Keep humans accountable for outputs. Set expectations and policies on review and quality.

  6. Begin adoption with an existing task. Involve the people who do the work in deciding what to automate.

  7. Give builders a governance path that is not punishing. Include the transformation lead in AI governance and treat key-person dependency as a risk.

  8. Protect the talent pipeline. Plan deliberately for entry-level roles and judgement.

Trusted Change Strategies Straight to your Inbox

Trusted Change Strategies Straight to your Inbox

Tap into the latest insights and proven success blueprints from change leaders around the world.

Tap into the latest insights and proven success blueprints from change leaders around the world.

ECLC logo

Meet Your Peers,
Join the Action.

ECLC logo

ECLC Conference 2026

The AI Summit for
Transformation Leaders

October 8, 2026, 7:30AM ET
City Winery, New York City

How Well is Your Org Activating Change?