When AI Doesn't Expand a Market – It Creates One From Scratch

01 /The Concept

CaseCraft says it knows the future – but only in one specific domain: how minor legal disputes will be handled in the UK.

The startup defines "minor" as civil cases with liability under £10,000, and personal injury or housing disrepair cases under £1,000.

To handle these, CaseCraft built a platform anchored by an AI model specifically trained on this category of cases.

Pricing is simple: £15 to file a claim, plus 10% of any award actually recovered. The platform supports both filing claims and defending against them.

The workflow is straightforward. A user logs in, uploads relevant documents, and the AI analyzes them and helps draft the appropriate submission. From there, the AI manages a calendar of required actions – court appearances, document deadlines – through to resolution.

CaseCraft is currently in closed beta with 100 initial users and has no full-time employees yet, only freelancers. Despite that, it has already raised its first £550,000 (~$742,000) in seed funding.

02 /Why It Matters

CaseCraft's founders run their own law firm. The startup idea came directly from their own experience: clients routinely approach the firm with exactly these kinds of small claims. But when they hear what legal fees would run, they walk away – because the potential award barely covers the cost, if it's awarded at all.

In other words, pursuing these cases through a law firm makes almost no financial sense for anyone involved. So neither side does. The potential claimants don't pursue it. The firms don't take it. Yet the market is real.

CaseCraft's goal is to unlock that market through AI – making the process cheap enough that it finally becomes worth doing. Those first 100 beta users, incidentally, are people who originally approached the founders' law firm with these exact claims before being redirected to the new platform.

The insight maps cleanly onto a broader pattern: breakthrough AI startups tend to emerge where they can reduce cost or complexity by 10–100x, enabling 10–100x more usage. CaseCraft is a more extreme version – it's not about getting more people to do something; it's about getting people to do it at all, because right now, essentially no one does.

This model is proving out across different corners of the legal market.

UK-based Valla ([covered here](/review/idti-nuzhno-tuda-chto-mozhet-rasshiritsja)) raised £2 million in June for a similar AI platform focused on employment rights.

Australian startup Ajust ([covered here](/review/udivitelnyj-sposob-zavoevat-ljubov-klienta)) raised $2 million for an AI consumer complaints platform – letting users file grievances against airlines for delayed flights or lost baggage, retailers who overcharged them, telecoms providing substandard service, and so on.

US-based Aparti ([covered here](/review/razvod-kak-servis)) built an AI platform specifically for filing divorce proceedings, with the option to escalate to a human attorney if property division or custody disputes get complicated.

And Pap! ([covered here](/review/zavlech-sobrat-i-dvazhdy-zarabotat)) raised $4.4 million for an AI agent that automatically claims price-match refunds. When a retailer drops the price of something you recently bought, you're typically owed the difference – but almost no one bothers to file for it. Pap!'s AI monitors receipt emails and price announcements, files the refund automatically, and takes a 20% cut of whatever it recovers.

03 /Opportunities

The broad direction is finding niches where AI can reduce complexity or cost by 10–100x, unlocking demand that simply didn't exist before.

But inside that broad direction, there's a sharper version of the thesis: find markets where virtually nothing is happening – not because there's no demand, but because it's too expensive or cumbersome to serve. Then make it cheap enough and simple enough that people actually start doing it.

The practical entry point: look at the customers you currently turn away – too small to be profitable, cases too simple to justify fees. That rejection list is a product roadmap. CaseCraft literally built its first 100-user cohort from clients who got turned away from the founders' own law firm.