Watched something interesting unfold at a client site last month. They’d spent $80,000 on an AI training program from a well-known vendor. Completion rates were good. Satisfaction scores were acceptable.

Meanwhile, a group of three employees had started an informal Slack channel sharing AI prompts and techniques. Within weeks, it had 60 members. People were posting use cases, troubleshooting together, celebrating wins.

The Slack channel was producing more actual behaviour change than the $80,000 program.

Why Peer Learning Works for AI

AI skills are different from traditional technical training. There’s no fixed curriculum because the technology changes monthly. Best practices emerge from experimentation. And the “right” approach depends heavily on context—your industry, your data, your workflows.

Peer learning fits this landscape naturally. Someone discovers a useful technique. They share it. Others try it, adapt it, improve it. Knowledge flows horizontally rather than trickling down from experts.

The people teaching are also the people doing the work. They know the real constraints—the crappy data, the legacy systems, the stakeholder expectations. Their solutions account for reality in ways vendor content can’t.

What Formal Training Misses

Formal AI training programs typically teach general concepts and generic examples. They have to—they need to apply broadly to justify development costs.

But the hard part of AI adoption isn’t understanding what’s possible. It’s figuring out how it applies to your specific situation. How do I get this tool to work with our unusual file formats? What prompts actually work for our type of customer enquiry? How do I explain this output to my sceptical manager?

Peers answer these questions. Formal training doesn’t.

Structuring Peer Learning

The most effective peer learning isn’t completely unstructured. It benefits from scaffolding that L&D can provide.

Visible channels. Create dedicated spaces—Slack channels, Teams groups, regular meetups—where AI experimentation is expected and encouraged. Make them easy to find for anyone curious.

Recognition. Acknowledge people who contribute useful knowledge. Doesn’t need to be formal rewards—public appreciation in all-hands meetings or newsletters works. What matters is signalling that sharing is valued.

Time allocation. Give people permission to spend time on peer learning. If it’s something they have to hide or do after hours, adoption will be limited.

Light facilitation. Someone should monitor the channel and occasionally prompt discussion. “Has anyone tried using AI for X?” can surface dormant expertise. But heavy-handed moderation kills organic exchange.

Curated highlights. Periodically surface the best tips, techniques, and use cases from peer channels. Not everyone reads everything. A monthly summary makes knowledge accessible to the broader organisation.

The Power of Local Examples

Here’s what I’ve noticed consistently: employees pay more attention to examples from their own organisation than external case studies.

When someone in accounting shares how they cut month-end reconciliation time by 30% using AI, the rest of the accounting team listens. When a training vendor shares a generic case study about “a financial services company,” eyes glaze over.

Local examples carry credibility. People know the context. They can ask questions. They can see themselves doing something similar.

L&D should actively collect and amplify these internal success stories. They’re worth more than purchased content.

Where Formal Training Still Matters

Peer learning isn’t a complete replacement. It works best for practical application once people have foundational understanding.

Formal training still makes sense for:

• Compliance and governance requirements (data handling, acceptable use)
• Advanced technical skills for specialist roles
• Onboarding new employees who need baseline knowledge
• Structured skill assessment and certification

Think of formal training as creating the foundation. Peer learning builds the house on top.

L&D’s Role in Peer-Driven Development

Some L&D teams feel threatened by peer learning. If employees teach each other, what’s our role?

The role shifts from content delivery to learning ecosystem design. You’re creating conditions for effective peer exchange, not controlling all the information flow.

This includes:

• Identifying and supporting informal knowledge leaders
• Connecting silos (different departments often solve similar problems independently)
• Filling gaps where peer expertise doesn’t exist
• Quality assurance on critical compliance topics
• Documentation and knowledge management

It’s a different skill set than traditional instructional design, but arguably more valuable for topics where knowledge evolves rapidly.

Getting Started

If your organisation hasn’t invested in peer learning infrastructure for AI skills, here’s a simple starting point:

1. Create a visible, optional AI experimentation channel
2. Seed it with a few enthusiastic early contributors
3. Post weekly prompts or challenges to stimulate discussion
4. Recognise valuable contributions in visible forums
5. Summarise highlights monthly for broader distribution

Cost: minimal. Impact: potentially significant.

The learning that matters often happens between the formal programs. L&D’s job is to make that informal learning easier, not to replace it with more courses.