Square launched a new ChatGPT app and a Claude plugin on July 1, 2026, letting consumers discover restaurants and place orders directly within those AI chat interfaces without leaving the conversation, VentureBeat reported the same day. The integration is live for any US-based Food & Beverage seller with an activated Square Online Ordering profile, with eligible merchants automatically opted in -- no new APIs to build, and no additional setup required on the restaurant's part.
The economics are the real story: rather than surrendering a 25-30% commission to a delivery aggregator, a restaurant discovered and ordered from through ChatGPT or Claude pays only Square's standard online transaction processing fee, which typically runs around 2.9% plus 30 cents per transaction on a standard plan, with no additional marketplace commission layered on top. For an independent restaurant that might clear only 3% to 9% net profit on a good day, handing over a 25-30% cut on a $40 digital order effectively means preparing that food at a loss -- making Square's low-fee approach a materially different economic proposition than the aggregator model most restaurants have been stuck with.
The mechanics are designed to feel invisible to the end consumer: when someone prompts ChatGPT or Claude with a request like finding a specialty coffee shop nearby with a great pour-over and ordering a bag of house roast, the AI parses real-time inventory and menu data Square provides, lets the customer browse results and make selections, then finalizes the purchase using Order by Cash App -- all without the user ever leaving the chat interface.
The move positions Square, rather than either AI lab, as the commerce and payments layer that actually captures the transaction volume flowing through agentic AI interactions -- a distinct and arguably more durable position than competing to build a checkout experience directly inside a chatbot. It's a meaningfully different bet than OpenAI or Anthropic building native commerce features themselves, instead treating Square's existing merchant and payments infrastructure as the plumbing underneath agent-driven discovery and ordering.
For restaurant owners and small-business operators, this integration is a concrete, immediately usable example of 'agentic commerce' delivering real bottom-line value rather than remaining a pilot or a conference-demo concept -- and given the automatic opt-in for existing Square merchants, adoption doesn't require any deliberate technical investment on the restaurant's part. For investors evaluating the broader AI-commerce landscape, Square capturing this distribution channel is a signal that payments and point-of-sale infrastructure incumbents may be better positioned to monetize agentic AI traffic than either the AI labs or the delivery aggregators they're partially disintermediating.
What to watch: how quickly consumer usage of ChatGPT- and Claude-initiated restaurant orders scales relative to traditional search and delivery-app discovery, whether delivery aggregators respond with their own fee restructuring to compete, and whether Square extends the same low-fee, AI-agent ordering model beyond food and beverage into other retail categories on its platform.