Summary Bullets:

• Google will release a broadly backed commerce protocol this quarter that threatens Amazon’s behemoth e-commerce marketplace

• Anthropic Claude Code includes new features that help developers bypass the complexities of COBOL, threatening IBM’s massive mainframe business.

The cloud wars have quickly morphed into the AI wars among leading platform services rivals, which have invested heavily in LLMs in recent years. By leveraging agentic AI, platform services providers are elevating their competitive threat in their quest for dominance in both e-commerce markets and among enterprise developer communities.

GlobalData’s GenAI Quarterly report released this week (Please see Generative AI Quarterly Q1 2026: Google Goes After Amazon’s Commerce Dominance), highlights aggressive moves among cloud competitors which seriously target arch-rivals’ deeply entrenched revenue sources. Earlier this quarter Google announced an agentic Universal Commerce Protocol (UCP), to eliminate the need for retailers to maintain complex infrastructures necessary to shore up their online purchasing lifecycles. This common language works across agents and systems to standardize most of the purchase journey including authentication, payment, and fulfillment. The alternative has been for DevOps team members to painstakingly integrate connections between various systems and partners, a major roadblock in efforts to digitize companies. Professional services required to connect infrastructure and modern applications are expensive and not always readily available. Google’s quest for a de facto agentic commerce standard is a major blow to its hyperscaler rivals, as UCP reduces the need for bespoke integration work, ultimately shifting leverage from competing platform services. It also addresses commerce partners’ strong desire to optimize AI to help drive new revenues.

In the same quarter, Anthropic revealed to its broad developer base new features within Claude Code that can modernize the legacy programming language COBOL. The language was created in 1959 but remains in use, shoring up multiple government and banking transactional processing systems, including IBM mainframes. Refactoring legacy code like COBOL represents a daunting task even among the industry’s most skilled coders. If Claude Code can significantly cut that learning curve via GenAI automation, the features represent a major new app modernization opportunity for enterprise developers to chase.