AI Agents Undermine Business Models of Amazon, Uber, and DoorDash – Analysts View Blocks as Justified

Platforms like Amazon, Uber and DoorDash face a new challenge: third-party artificial intelligence agents could in future mediate directly between consumers and providers, thereby undermining the traditional business model of aggregators. Analysts at the investment bank Bernstein currently view the technical and legal barriers that some companies have already erected as justified.

The Battle for Customer Access

At its core, the question is who will control direct access to the customer in future. External AI agents could function as digital assistants that, on behalf of users, compare prices, make bookings or place orders – without the customer themselves visiting the platforms of Amazon, Uber or DoorDash. The aggregators would thus be relegated to interchangeable suppliers, while the AI agents occupy the interface with the end customer.

This constellation is not new. As early as November 2025, the technology magazine The Verge analysed the central question "Who owns the customer?" for service companies in the app era, including Uber, Lyft, Airbnb and others. Industry observers at the time reckoned that companies would attempt to block AI agents entirely rather than cede customers to these intermediaries.

DoorDash Particularly Vulnerable

The food delivery service DoorDash appears particularly vulnerable. According to an analysis by Citrini Research from 16 October 2026, AI agents could bypass delivery services entirely and conduct business directly with restaurants. Commissions would then be around 7 percent – significantly below the usual intermediary fees that DoorDash charges. In this scenario, the AI agents would independently contact restaurants, place orders and possibly also organize delivery, without DoorDash being needed as a middleman.

Technical Barriers Already in Use

Several large companies have already taken countermeasures. According to a report by aigovernance.com from 17 October 2026, Amazon, Delta Air Lines, JPMorgan Chase and Charles Schwab rely on technical controls, contractual restrictions and legal threats to limit AI agents' access to their services. Technical measures range from detecting and blocking automated access to changes in terms of use that prohibit the use of third-party software.

Analysts' Rationale

Bernstein argues that the blocks are at least understandable for now, as the AI agents directly attack the business model of aggregators. The platforms have made substantial investments in technology, logistics and customer loyalty over many years. Amazon built a comprehensive logistics network, Uber and DoorDash established driver and delivery partner networks. External AI agents could use this infrastructure without making comparable investments themselves.

The investment bank sees the threat as greatest where aggregators are particularly well positioned: in bundling offerings and simplifying the selection process for consumers. Exactly this function could be assumed by AI agents – but without the commissions and advertising revenues that form the core of today's revenue models.

Long-term Perspective Unclear

How durable the current blocks can be remains open. Regulatory authorities could step in if platforms use their dominant market position to systematically exclude competitors. Moreover, AI agent technology is advancing rapidly. Should they establish themselves as a standard interface between consumers and digital services, platforms may be forced to adapt their strategy – or lose market share.

The dispute between established aggregators and emerging AI intermediaries is likely to characterise the coming years. For investors, the question arises which business models will have long-term viability in this environment and which companies are flexible enough to adapt to changed customer interfaces.