Last Thursday’s CLE webinar on AI governance drew 312 in-house counsel — our largest live audience ever. Rachel Kim, GC at CloudSync, walked through the one-page AI policy framework she developed, and the Q&A ran 25 minutes past schedule because nobody wanted to leave.
For those who couldn’t attend, here are the five takeaways that resonated most — based on the live poll responses and the questions that came through the chat.
1. Most companies are using AI without a policy
We polled attendees at the start: “Does your company have a formal AI acceptable use policy?” The results were stark: 14% said yes, 38% said AI is being used but there’s no policy yet, 29% said “not officially, but people are using it anyway,” and 19% said they’ve actively restricted AI or haven’t addressed it. That means roughly two-thirds of the companies represented are in some form of “yellow zone” — AI is happening, legal knows it, but there’s no framework governing it.
2. A one-page policy beats a 40-page policy
The most downloaded resource from the webinar was the one-page AI Acceptable Use Policy template. Rachel’s core insight: complexity is the enemy of adoption. If your policy requires a 30-minute training session to understand, nobody will follow it. The power of the “traffic light” framework — green light, yellow light, red light — is that employees can self-sort 80% of their AI use cases without needing to ask legal.
3. Build in a sunset clause
AI changes fast. A policy drafted in January might be outdated by April. Rachel’s recommendation: include a 90-day review date on every version of your policy. Put it in your calendar. When the date comes, spend 30 minutes asking: “Has anything changed that makes this policy wrong?”
4. The question nobody could answer
The Q&A session was robust, but one question stumped everyone — including Rachel: “What do you do when your CEO has been using ChatGPT to draft board communications for six months and you just found out?” The honest answer: there’s no playbook for this. That’s the reality of working in AI governance in 2026. We’re all figuring it out together.
What’s next
Based on the demand from this session, we’re planning a follow-up webinar on vendor AI contract negotiation — the 12 questions every GC should ask before signing an enterprise AI agreement. That’ll be in April. If you want to be notified when registration opens, make sure you’re subscribed to the IHC newsletter.
