Legal AI startup Sandstone's approach shows that the model is a commodity. Real defensibility comes from creating a "context layer" that integrates data from CRM, CLM, and communications, giving the AI the business context required to be truly useful for in-house teams.

Crosby's business model is to be an AI-powered law firm, selling end-to-end legal work rather than a software tool. This allows them to fully leverage automation and capture the entire value of the work performed, a more defensible strategy than selling a legal copilot that competes with foundation models.

While foundational AI models threaten broad applications like writing aids, startups can thrive by focusing on vertical-specific needs. Building for niche workflows, compliance, and deep integrations creates a moat that large, generalist AI companies are unlikely to cross.

A key competitive advantage for AI companies lies in capturing proprietary outcomes data by owning a customer's end-to-end workflow. This data, such as which legal cases are won or lost, is not publicly available. It creates a powerful feedback loop where the AI gets smarter at predicting valuable outcomes, a moat that general models cannot replicate.

Instead of selling AI co-pilots, legal tech startup Crosby operates as a full-stack law firm using AI internally. This model allows them to continuously re-orchestrate workflows between human lawyers and AI as models improve. This captures the entire value of automation rather than just the limited margin from selling a software tool to other firms.

While many legal AI tools use the same foundational models, they differentiate by offering features crucial for law firms: strict permissions, compliance controls, and integrations with proprietary legal databases like Westlaw. This 'packaging' of trust is the real product, for which discerning law firms willingly pay a premium.

Companies create defensibility by generating unique, non-public data through their operations (e.g., legal case outcomes). This proprietary data improves their own models, creating a feedback loop and a compounding advantage that large, generalist labs like OpenAI cannot replicate.

YC Partner Harsh Taggar suggests a durable competitive moat for startups exists in niche, B2B verticals like auditing or insurance. The top engineering talent at large labs like OpenAI or Anthropic are unlikely to be passionate about building these specific applications, leaving the market open for focused startups.

In the AI era, defensibility comes from building a complex system of record, not just a thin wrapper on an LLM. Companies with a 'thick application layer' that offers standalone value are unattractive for model providers to replicate, whereas thin wrappers risk being absorbed by the platform they are built on.