Alphabet's CEO, Sundar Pichai, addressed attendees at the AI Impact Summit in New Delhi on February 20, 2026, showcasing the company's latest advancements. Following a protracted period of losses on Wall Street that lasted over ten years, Google has ushered in September with renewed optimism for its investors. On Wednesday, the tech giant introduced Gemini 3.8 Flash—its third Flash model in just six weeks—alongside a new cybersecurity model designed for trusted government and enterprise clients. Berkshire Hathaway's CEO, Greg Abel, publicly affirmed confidence in Alphabet’s position in the AI space. In further favorable news, a federal judge dismissed the Justice Department’s attempt to compel Google to divest its ad exchange, marking another success in the company’s battle against government antitrust initiatives. These developments collectively offer a more promising outlook for Alphabet after a challenging summer that saw a slowdown in its AI advancements, notable talent exits, and significant restructuring within DeepMind.
While the stock experienced a slight uptick of 0.6% on Wednesday, it still faced a minor decline after dropping over 1% the previous day. Google remains optimistic about Gemini 3.8 Flash, which emphasizes coding and agentic tasks—areas in which AI companies are striving to translate technological progress into enterprise revenue. Google deems this latest version as the most advanced in reasoning and coding, boasting substantial enhancements over its predecessor, 3.7 Flash, particularly in software engineering and multi-step tasks. According to Tulsee Doshi, senior director of product management at Google DeepMind, the recent Flash models have exceeded expectations, providing opportunities for deeper integration into their offerings.
The smaller Flash models are not only more cost-effective to run but have also shown rapid iteration capabilities compared to larger systems, although they have yet to catch up with competitors like Anthropic and OpenAI. Analyst Gil Luria from D.A. Davidson stated that while this product helps Google maintain competitiveness, it likely won’t change its status as a distant third in the enterprise market, recommending investors hold their positions.
Pricing plays a crucial role in Google’s strategy. Gemini 3.8 Flash is available at $0.75 per million input tokens and $3.75 per million output tokens, maintaining the same introductory price as the previous Flash model, despite its improvements. Gemini Enterprise is also launching pay-as-you-go pricing options, token discounts of up to 20%, monthly caps on spending, and a zero-cost base subscription. According to reports shared with CNBC, Google is actively challenging Microsoft and Anthropic, asserting that their recurring fees and separate product licenses render them pricier and less adaptable. Furthermore, Google Cloud enjoys significant scale, with approximately 75% of its customers already utilizing its AI products. Google Cloud CEO Thomas Kurian noted that these clients are spending about 50% more than originally agreed upon. Demis Hassabis of DeepMind discussed a future strategy where Gemini serves as a universal framework coordinating more affordable, specialized models and agents, emphasizing that overall system breadth may be as critical as having the most sophisticated model.
Additionally, Google highlighted the cost advantages in cybersecurity, stating that Gemini 3.8 Flash Cyber can identify and resolve software vulnerabilities with performance comparable to leading models while operating much faster and at a lower cost. Doshi expressed excitement about providing a cybersecurity solution that offers superior performance at a fraction of the cost, although access will initially be restricted to a select group of trusted government and enterprise defenders through the Fairwind Program. This strategy involves a tradeoff, as Alphabet is investing heavily in AI infrastructure, banking on growth and market share increases to yield long-term returns.
Berkshire's Abel elaborated to CNBC about viewing Alphabet as an AI frontrunner, based on insights gained from how its portfolio companies benefit from the technology. He stated that the visibility gained from within these companies prompted increased interest in Google as a significant player in AI. Despite some market analysts perceiving Google’s AI position as precarious, the advertising segment remains robust, registering a 14% growth in the last quarter. The ad business received another boost this week when a federal judge ruled that Google would not be required to divest its AdX exchange, selecting behavioral solutions over the structural breakup regulators had aimed for. This decision succeeded another antitrust ruling from the previous year, in which a judge denied calls to separate Google from Chrome. Antitrust attorney Wyatt Fore remarked on the significance of both rulings, indicating that they empower Google as it progresses in the AI sector without constraints holding it back.
