AI Champions
AI Champions are network-identified individuals whose structural position within the organisation gives them the social capital to drive organic, peer-to-peer AI adoption — far more effectively than top-down mandates alone.
The problem with top-down AI adoption
Most organisations attempt AI adoption through executive mandates: a leadership decision is made, a strategy is announced, training programmes are rolled out, and employees are expected to change their behaviour. The results are predictably mixed. Survey after survey shows that AI adoption rates stagnate not because tools are unavailable, but because social trust is missing.
Employees adopt new technologies when people they already trust model and advocate for those technologies. The challenge is identifying who those trusted individuals are — because they rarely appear on the org chart. This is precisely the problem that ONA and AI Champion programs are designed to solve.
The three AI Champion archetypes
Using betweenness centrality and Infomap community detection applied to multiplex ONA data, my research has identified three distinct archetypes that every effective AI champion program needs:
Oracle
High cross-layer betweenness centrality
Oracles are the most central individuals across multiple network layers — the people everyone turns to for advice, information, and trust. Their AI endorsement carries maximum credibility and propagates rapidly through the network. When an Oracle says "this AI tool changed how I work," the message is heard and believed.
Program role: Core advocates and visible sponsors of the AI champion program. Prioritise for early AI tool access, deep training, and public recognition.
Broker
High inter-community betweenness
Brokers sit at the boundary between organisational communities (departments, divisions, or informal groups identified by Infomap). They are the translators and connectors who relay knowledge across silos. A Broker in Sales who is also connected to Engineering can transfer AI use-cases between otherwise disconnected teams.
Program role: Cross-functional AI knowledge transfer agents. Equip Brokers with AI case studies from multiple domains and empower them to facilitate cross-departmental AI learning sessions.
Silo Buster
Low internal centrality, high inter-community ties
Silo Busters are not the most prominent figures within their own teams, but they maintain meaningful connections across multiple organisational communities. Without deliberate program design, isolated teams risk developing divergent AI practices or — worse — remaining entirely un-reached by AI adoption efforts. Silo Busters prevent this by acting as the last-mile connectors.
Program role: Outreach specialists for hard-to-reach pockets of the organisation. Identify which isolated communities they bridge and task them with ensuring no team is left behind.
Designing an AI Champion Program
An effective AI Champion Program built on ONA data follows five phases:
- ONA diagnostic — administer a multiplex ONA survey to map all five network layers across the target population.
- Archetype classification — compute betweenness centrality per layer, run Infomap community detection, classify each individual as Oracle, Broker, or Silo Buster.
- Program design — define role-specific Champion responsibilities, training content, time allocations ("champion hours"), recognition mechanisms, and success metrics.
- Activation — onboard Champions with dedicated AI training, communication templates, and access to an internal Champion community of practice.
- Measurement — re-administer the ONA after 6–12 months to track whether the AI Support network has expanded, silos have dissolved, and adoption rates have improved.
Frequently asked questions
What is an AI Champion?
An AI Champion is an employee identified through ONA as having the network position and social capital to credibly advocate for AI adoption among peers. Unlike top-down executive mandates, AI Champions influence adoption organically through trusted relationships.
What are the three AI Champion archetypes?
Oracle: high betweenness centrality across all network layers — the expert everyone trusts. Broker: high betweenness between community clusters — the cross-boundary connector. Silo Buster: low internal centrality but high inter-community ties — prevents knowledge hoarding and spreads AI practices beyond isolated teams.
How are AI Champions identified using ONA?
We administer a multiplex ONA survey to map five network layers. Betweenness centrality is computed per layer and aggregated. Infomap community detection identifies cluster boundaries. Individuals are then classified as Oracle, Broker, or Silo Buster based on their structural position.
Why are AI Champions more effective than top-down AI mandates?
Change research consistently shows that peer influence is more persuasive than authority. AI Champions are already trusted within their social networks — making their AI endorsements far more credible and behaviour-changing than directives from senior leadership.
How is an AI Champion Program designed?
After ONA-based identification, Champions receive specialised training, a mandate to share learnings, structured "champion hours" to assist colleagues, and periodic recognition. ONA is re-administered after 6–12 months to measure network change.