Alphabet and Google CEO Sundar Pichai used the company’s latest earnings call to defend Google’s position in frontier AI after investor questions intensified around the delayed release of Gemini 3.5 Pro and the company’s competitive standing in AI coding and agentic workflows. The questions followed concerns that Google had lost momentum at a time when OpenAI, Anthropic and Chinese AI companies have accelerated model releases. Pichai told analysts that Google continues to have frontier models, while acknowledging that coding and agentic coding are areas where improvement is needed.
The exchange is significant for enterprise technology buyers because model leadership is no longer an abstract benchmark contest. Indian IT services firms, GCCs, digital-native businesses and regulated enterprises are increasingly selecting AI platforms for software engineering, cybersecurity, customer-service automation, data analytics and business-process transformation. Pichai’s emphasis on Gemini Flash, described as Google’s cheaper and faster workhorse model, points to an enterprise AI strategy built around production efficiency rather than only flagship-model performance. Google has positioned Gemini Flash for use cases including cybersecurity, customer-service agents, data analytics and enterprise software.
The delayed Gemini 3.5 Pro release has put additional scrutiny on Google’s AI roadmap, particularly because coding assistants and agentic software-development workflows are becoming strategic battlegrounds. For India’s technology services ecosystem, where software engineering productivity, application-modernisation work and AI-assisted delivery models are central to competitiveness, the relative strength of Google’s models against other leading systems has direct implications. If enterprises perceive that Google is lagging in coding or autonomous task execution, that could influence platform choices across cloud, developer tooling and AI productivity suites.
Pichai’s remarks also show how AI competition is moving from research performance to commercial deployment economics. Many Indian enterprises are already evaluating cost, latency, security posture, data controls and integration depth when selecting models. A lower-cost, faster model that is reliable across high-volume tasks may be more useful for production workloads than a larger model reserved for frontier reasoning. Google’s defence therefore appears focused on showing that its portfolio can serve both advanced AI development and day-to-day enterprise adoption.
The leadership moment also carries a broader CXO signal: technology buyers should expect the frontier AI market to remain volatile, with roadmaps, model launches and pricing strategies shifting quickly. Google’s response suggests the company is trying to reassure customers and investors that temporary launch delays do not weaken its long-term platform position. For India-facing organisations standardising on AI infrastructure, the next phase will likely hinge on whether vendors can translate model claims into dependable workflow automation, developer productivity and governance-ready deployments.
