Shopify Abandons React Native for Swift & Kotlin

// Swift implementation func processOrder(order: Order) async throws -> OrderResult {
 let validated = try await validateOrder(order)
 return try await fulfillOrder(validated)
}

// Kotlin implementation suspend fun processOrder(order: Order): OrderResult {
 val validated = validateOrder(order)
 return fulfillOrder(validated)
}

Shopify announced in September 2026 that it’s abandoning React Native in favor of maintaining separate Swift and Kotlin codebases for iOS and Android. The decision reverses a 2020 strategy shift and highlights how AI coding agents have changed the economics of mobile development.

Key Findings

According to Shopify’s announcement, the company originally adopted React Native in 2020 for three specific reasons: eliminating duplicate feature development across platforms, enabling developers to work across the entire technology stack, and reducing time spent on feature parity in favor of shipping new functionality.

Six years later, Shopify concluded that while native development still requires building and maintaining software on two separate platforms, AI agents have fundamentally altered the cost calculation. The company stated that agents can now handle enough of the implementation, translation, testing, and review work that the cross-platform benefits of React Native no longer justify its use as the deciding factor.

The announcement acknowledged React Native as a strong platform during the six years Shopify used it. The company maintains three significant React Native libraries: react-native-skia, flash-list, and restyle. The first two libraries will continue under new stewardship, while restyle will be archived at the end of 2026 due to having a smaller user base compared to Shopify’s other open-source projects.

Methodology

Shopify’s decision appears driven by a direct comparison between the 2020 development landscape and current capabilities. In 2020, maintaining parallel Swift and Kotlin codebases meant developers had to manually implement each feature twice, coordinate releases across platforms, and constantly monitor for diverging functionality.

The shift back to native development suggests that AI coding agents now provide sufficient automation for tasks like translating features between Swift and Kotlin, generating platform-specific tests, and conducting code reviews. This automation apparently reduces the labor overhead enough to make native development competitive with React Native’s write-once approach.

Implications

The reversal carries implications beyond Shopify’s mobile strategy. It represents one of the first major architectural decisions by a significant technology company explicitly justified by AI agent capabilities. Rather than treating coding agents as productivity enhancers within an existing architecture, Shopify used them as grounds to fundamentally restructure its mobile development approach.

For the React Native ecosystem, Shopify’s departure removes a major contributor. The company’s three libraries served different purposes in the React Native community, and while react-native-skia and flash-list will find new maintainers, the transition period may create uncertainty for projects depending on these tools.

The decision also raises questions about other cross-platform frameworks. If AI agents can economically support parallel native codebases, similar reasoning might apply to web frameworks, desktop applications, or backend services currently unified under single technologies for developer efficiency.

Bottom Line

Shopify’s move from React Native back to native development marks a concrete example of AI coding agents influencing architectural decisions at scale. The company determined that the implementation, translation, testing, and review work agents can now perform has shifted the cost-benefit analysis that originally drove its 2020 adoption of React Native.

Whether this represents a broader trend depends on how other organizations evaluate similar tradeoffs. Shopify’s experience suggests that for teams with sufficient scale and AI tooling maturity, the performance and platform-specific advantages of native development may now outweigh the coordination costs that previously made cross-platform frameworks attractive.