Alphabet CEO Sundar Pichai strongly defended Google's artificial intelligence (AI) strategy during Wednesday's earnings call, countering concerns that the company is falling behind due to delays in its flagship model and perceived lags in AI programming.
Investor unease has grown due to the postponement of Gemini 3.5 Pro, a model initially slated for a June release that was highly anticipated to boost Google's competitiveness in AI programming and autonomous "agent" tasks—areas currently at the forefront of industry competition.
The latest delay has intensified market worries that Google may be losing momentum as competitors like OpenAI, Anthropic, and numerous Chinese rivals accelerate their own model release cycles. During the call, when analysts pressed on the status of Google's frontier models and its ability to remain competitive at the cutting edge of AI research, Pichai adopted a notably more defensive stance than usual.
Addressing questions from JPMorgan analyst Doug Anmuth regarding Gemini's ability to maintain a leading competitive edge and Google's slower release pace, Pichai stated, "We clearly have had frontier models all along. In many ways, we are still at the frontier; there are areas we acknowledge we need to improve, and programming and agent programming are among them."
Pichai did not elaborate extensively on the delay of Gemini 3.5 Pro, instead repeatedly highlighting Gemini Flash—Google's lower-cost, faster-performing "workhorse" model. This model is already being widely deployed in an expanding range of applications, including cybersecurity, customer service agents, data analytics, and enterprise software.
He specifically pointed to this week's release of Gemini 3.6 Flash, noting it shows a more than 10-percentage-point improvement in coding benchmarks over the prior version while using fewer tokens. Google also released Gemini 3.5 Flash-Lite and the Flash Cyber model for cybersecurity on Tuesday, while Gemini 3.5 Pro remains in testing with partners.
"I think when we ship Gemini 4, people will be happy," Pichai said, describing it as "an ambitious effort." He emphasized that Google is training a significantly larger model with the goal of competing at the technical frontier upon its release, adding that the company remains "very determined and confident" in maintaining its leadership in AI.
When Barclays analyst Ross Sandler raised similar concerns about Google's slower model release cadence, Pichai revealed that the product roadmap for Gemini 4 plans for new model releases at a frequency of "almost once a month."
Pichai's robust defense underscores an increasingly critical challenge for Google: convincing investors to judge its AI ambitions not solely on the delay of a single flagship model, but rather on its vast AI ecosystem. This system encompasses cloud infrastructure, custom AI chips, and a matrix of Gemini models embedded across Google's diverse products and services.
Shares of Alphabet (GOOGL) fell more than 3% in after-hours trading. Concerns over Gemini project delays, coupled with the departure of several executives, have contributed to a stock decline of approximately 9% since late April.
Comments