Has AI Killed Product Moats?

In the age of AI, defensibility looks different

I recently caught up with a founder friend and we got talking about defensibility. With AI changing the pace of new software, how fast are moats shrinking? How much will product lifespans be cut back? Will being a first mover matter when fast followers can build a competitive offering in a matter of months (or weeks)?

When I started venture investing ~10 years ago, we talked a lot about intellectual property. At the time, IP = defensibility and was part of almost every new company assessment.

But what is IP in the age of AI?

It’s hard to imagine that agents won’t continue to take on more of our work. Does the human conductor have rights to the AI’s creation? Will we focus on the data used to create agents instead?

Even today, the training data argument feels weak. Despite copyright lawsuits, ChatGPT has 500M weekly users. And pre-training may no longer matter as the internet - the most significant digital data source we have - has already been plugged into today’s models.

Apart from IP, many functions will fundamentally change. Products - and entire operations - may not be needed in the future.

Will learning & development departments exist when employees can learn everything just-in-time? Will agents market and sell to one another, eliminating humans? Will security become agent attackers vs. agent defenders?

The pace of change is challenging everything. And what can be considered a true “moat” has also fundamentally changed.

For an AI-native world, moats come from 3 things - user experience, cadence, and momentum.

User experience is different for an AI product. The user needs more finesse and understanding of the product’s capabilities and nuances. And, at least for now, training or guardrails are required for a great user experience.

The quality of onboarding matters a lot - which is why we’re seeing a spike in roles for solutions and sales engineers.

And a thoughtful journey from first “wow” to expanded usage is very much needed. Too many AI products try to do too much, too fast. Users don’t want to spend hours figuring out how to get value from an AI product.

And, much like recommendation systems have a cold start problem, AI-driven SaaS has its own initial limitations. How will AI personalize to you without knowing you first?

This is why eval data is so important. Without that feedback loop, the chances of an exceptional user experience for an AI-driven product are low. There are many unknown unknowns that come with the real world.

The hands-down best user experience I’ve had with an AI product is ChatGPT. You easily onboard yourself, have an immediate wow after your first natural language interaction, and effortlessly expand your usage as it gets more personalized with every use.

In addition to user experience, and arguably the most important differentiator, is cadence. The pace of shipping, learning, and recruiting compounds. The very best teams I’ve worked with have always had this - but it’s a totally different ball game today.

Just this week, Harvey and Decagon announced fundraises at $5B and $1.5B after just a few years (or less) in market. And Harvey’s reported to have revenue in the $100M range. From what I’ve seen, both have recruited and shipped at a pace unseen at most SaaS predecessors. The bar has officially been raised.

Finally - momentum. I recently hosted an engineering leaders dinner and the conversation shifted to coding copilots. Despite the increasing number of products in market, the only 2 talked about were Windsurf and Cursor.

Why? They couldn’t point to specifics - just buzz. They *felt* like they were the ones that mattered.

A lot of that comes down to content, but fundraising, conferences, events, relationships, and happy vocal customers will all contribute. And the more unique the distribution strategy, the better.

Just look at Cluely - whether you like them or not, they’ve created insane buzz. My entire feed over the weekend was filled with their content - and many (many) takes from friends in tech. That mindshare will build on itself and be hard for another startup to come and displace.

In short, AI hasn’t killed moats—but it’s completely redefined them. Defensibility now lies in speed, user experience, and momentum. The game has changed; the winners will be those who move fast, adapt faster, and deeply understand their user journey.