Where Do AI Savings Go: Share Gains or EBITDA Expansion?

Most people agree, as do I, AI will meaningfully change cost structures across all businesses. Lower headcount, reduced software spend, faster cycle times. The more interesting question is what owners and investors actually do with that efficiency. If AI drives lower operating expense or higher throughput, that should translate to higher EBITDA assuming pricing holds. But whether that margin expansion is durable is a separate question. If these tools are broadly accessible, you have to assume some level of price compression over time as competitors catch up. In that case, underwriting full margin expansion as permanent is probably aggressive. For early adopters, there are a few paths. Hold pricing and expand margins, reinvest into growth, or pass savings through to customers to take share. The right answer likely depends on how defensible the implementation is. If it is embedded in process or tied to proprietary data, you may have a window of durable advantage. If not, it is likely transient. Curious how others are underwriting this today. Are you assuming AI-driven cost savings accrue to equity or get competed away? Would also love to hear what folks have actually seen so far or how you are thinking about it in practice.