Loading the player…

The surge in generative AI sparked rapid startup creation, but as the initial excitement fades, two previously popular business models are now emerging as potential warnings: LLM wrappers and AI aggregators. Darren Mowry, who heads Google’s global startup organization spanning Cloud, DeepMind, and Alphabet, suggests startups relying on these approaches have their “check engine light” illuminated.

LLM wrappers are startups that build a product or user experience layer around existing large language models—such as Claude, GPT, or Gemini—to address a particular need. An example is a company using AI to assist students with studying. “If you’re really just counting on the back end model to do all the work and you’re almost white-labeling that model, the industry doesn’t have a lot of patience for that anymore,” Mowry noted on a recent episode of Equity. He explained that wrapping “very thin intellectual property wrapped around Gemini or GPT-5” indicates a lack of differentiation. For a startup to “progress and grow,” he emphasized, “You’ve got to have deep, wide moats that are either horizontally differentiated or something really specific to a vertical market.”

Examples of LLM wrappers with substantial defensibility include Cursor, a GPT-powered coding assistant, and Harvey AI, a legal AI assistant. Essentially, the era when startups could simply overlay a user interface on a model like GPT and gain product traction—as might have been possible around mid-2024 with the launch of the ChatGPT store—has passed. The focus now is on constructing enduring product value.

AI aggregators represent a specific category of wrappers. These are startups that combine multiple LLMs into a single interface or API layer, directing queries across various models and providing users with access to several options. Typically, these platforms offer an orchestration layer that includes monitoring, governance, or evaluation tools. Examples include AI search startup Perplexity and developer platform OpenRouter, which supplies access to numerous AI models through one API.

Although many such platforms have gained users, Mowry offers straightforward advice to new startups: “Stay out of the aggregator business.” He observes that aggregators generally are not experiencing significant growth or advancement currently because users desire “some intellectual property built in.” This ensures queries are directed to the appropriate model at the right time based on user needs, rather than due to underlying computational or access limitations.

Mowry brings decades of experience in the cloud sector, having worked at AWS and Microsoft before joining Google Cloud, and he has witnessed similar cycles unfold. He draws a parallel between the present situation and the early days of cloud computing in the late 2000s and early 2010s, as Amazon’s cloud business began to accelerate. During that period, numerous startups emerged to resell AWS infrastructure, promoting themselves as simpler entry points that offered tooling, consolidated billing, and support. However, when Amazon developed its own enterprise tools and customers became adept at managing cloud services directly, most of those intermediaries were marginalized. Only those that added genuine services—such as security, migration, or DevOps consulting—survived.

Today, AI aggregators face comparable pressure on margins as model providers expand their own enterprise features, potentially bypassing middlemen. On a more optimistic note, Mowry is enthusiastic about vibe coding and developer platforms, which had a landmark year in 2025. Startups like Replit, Lovable, and Cursor—all Google Cloud customers, according to Mowry—attracted substantial investment and customer interest.

He also anticipates robust growth in direct-to-consumer technology, particularly companies that place powerful AI tools directly into customers’ hands. Mowry highlighted the opportunity for film and television students to use Google’s AI video generator, Veo, to bring their stories to life.

Looking beyond AI, Mowry believes biotech and climate tech are currently prominent fields. This is evident both in the venture investment flowing into these industries and in the “incredible amounts of data” available to startups, enabling them to generate real value “in ways we would never have been able to before.”