Alphabet shares slip as Gemini 3.5 Pro delay exposes AI gaps
Alphabet's flagship Gemini 3.5 Pro model is months behind schedule due to disappointing coding results, raising investor concerns about the return on the company's massive AI infrastructure spending.
Alphabet is running months behind schedule on the release of its flagship Gemini 3.5 Pro artificial intelligence model after an internal effort to improve its coding abilities fell short. The company had explicitly promised the upgrade for June, following the mid-May introduction of its less powerful sibling, Gemini 3.5 Flash.
The missed deadline has coincided with a notable pullback in the stock. Alphabet shares closed Friday at $346.77, down roughly 15% from a peak of $408.61 earlier this year. For a stock that has served as a primary market proxy for AI enthusiasm, this wobble reflects growing investor anxiety that the company is losing ground.
Google updated the data used to train Gemini 3.5 Pro specifically to enhance its coding performance, but the results proved disappointing. Coding is now arguably the main battleground for AI labs and a major driver of enterprise revenue. Sitting on a flagship model that cannot clear its own internal bar while rivals OpenAI and Anthropic continue to ship updates is an uncomfortable position for the search giant.
Alphabet pushed back against the narrative of a stalled pipeline. "We're currently testing 3.5 Pro, an upgraded Flash model, and other models with partners," the company said in a statement. The company added that it is "shipping quickly across a wide range of models while keeping them highly cost-effective for customers."
The ultimate significance of the delay hinges on Alphabet’s immense capital commitments. The tech giant expects capital expenditures to reach as much as $190 billion this year, with the vast majority funding the infrastructure behind its AI push. That level of spending demands a product portfolio that remains competitive at the technological frontier. A single missed deadline is manageable, but a pattern of slipping models would raise serious questions about the return on that investment.