Restaurants can now accept orders placed directly from ChatGPT and Claude thanks to Square's new, low-fee, no setup integration

Square is launching an Order-by-Cash App experience inside ChatGPT plus a Claude plugin, letting customers discover restaurants and place orders directly in chat. The key economic angle is that Square routes these AI orders into merchants’ existing POS flow while avoiding the typical 15%–30% marketplace commissions charged by DoorDash/Uber Eats—merchants instead pay Square’s standard online processing fees of about 2.9% + $0.30 (or 3.3% + $0.30 on certain plans). Square also positions this as part of a broader “agentic commerce” push, citing 42% of consumers using AI for shopping assistance and aiming for standards work such as Google’s Universal Commerce Protocol (UCP).

Analysis

This is mainly a routing fight, not a restaurant-tech story. If AI assistants become the default discovery layer, the economics migrate from marketplace toll collectors to whoever controls the merchant operating system, which is structurally better for SQNNY/Block and the AI-platform owners than for delivery intermediaries. The near-term P&L effect on SQNNY is probably modest because it still monetizes at standard payment rates, but the strategic value is higher: lower churn, more embedded order flow, and a cleaner path to capture incremental GMV without subsidizing demand.

For DASH and UBER, the risk is not a sudden revenue cliff; it is gradual margin pressure as commission rates become easier for merchants to benchmark and negotiate away. If AI surfaces siphon off high-intent pickup and repeat orders, the most profitable cohorts leave first, forcing the aggregators to defend share with lower take rates or more expensive merchant incentives. That tends to show up first in local merchant commentary and order-mix disclosures over the next 1-3 quarters, then in valuation as the market assigns a lower durable take-rate multiple.

The contrarian miss is adoption speed: AI discovery can be impressive in demos yet still account for a tiny share of transaction volume if consumers keep using search, maps, or native apps. The key falsifier is whether AI-origin orders become a measurable line item in merchant dashboards within 2-3 quarters; if not, the threat is more narrative than economic. Also watch whether GOOGL/Amazon own the customer interface and leave SQNNY as back-end plumbing, which would cap the upside to software economics rather than create a new platform premium.

Key Decisions for Investors

• Short DASH vs long SQNNY/Block for 1-3 months: best expression of fee-disintermediation risk; target a 10-15% relative move if AI-order routing gains traction, stop if DASH re-accelerates order growth or holds take-rate guidance.
• Long GOOGL vs short UBER as a 2-quarter pair: GOOGL gains intent/control of commerce surfaces, while UBER faces higher risk of channel dilution; risk/reward improves if Google starts bundling agentic checkout into Search/Gemini.
• Buy SQNNY on pullbacks only if merchant adoption metrics confirm AI-origin orders are becoming material; otherwise treat as a watch item, not a core long, because the near-term fee uplift is limited.
• Set a catalyst alert for next restaurant/merchant disclosures: if DASH or UBER management cites lower commission pressure or rising direct-order substitution, add to shorts; if not, fade the reaction as overdone.