The battle for the $30 trillion consumer market has begun! Alphabet Inc. Class C (GOOGL.US) launches Gemini AI agent, Morgan Stanley is bullish on its user scale advantage.

date
23:55 18/11/2026
avatar
GMT Eight
Google's parent company Alphabet has officially launched "Tasks," a new Gemini AI agent feature, upgrading the AI assistant from answering questions to a digital assistant capable of performing real-world tasks such as shopping, bookings, and payments on users' behalf.
The battle for the $30 trillion consumer market has begun! Alphabet Inc. Class C (GOOGL.US) launches Gemini AI agent, Morgan Stanley is bullish on its user scale advantage. Alphabet Inc. Class C's parent company Alphabet (GOOGL.US, GOOG.US) has officially launched a new Gemini artificial intelligence (AI) agent feature called "Tasks," upgrading the AI assistant from answering questions to a digital assistant capable of executing real tasks such as shopping, bookings, and payments on behalf of users. In a research report published on November 18, Morgan Stanley noted that Alphabet Inc. Class C, with Gemini's massive existing user base, deeply integrated application ecosystem, and significantly improved operating efficiency of its new-generation AI models, is expected to secure a favorable position in the increasingly fierce consumer AI agent competition in 2027. The bank believes that Alphabet Inc. Class C's latest move will further intensify competition with tech giants such as Meta (META.US) and Amazon.com, Inc. (AMZN.US). Compared with Meta Muse, which requires users to download separately, Gemini Tasks can leverage the existing user base of its apps for rapid promotion. Morgan Stanley maintains its "Overweight" rating on Alphabet and a $375 price target, implying approximately 34% upside from the November 17 closing price of $280.20. Gemini's new AI agent feature initially covers approximately 250 million users According to the research report, Alphabet Inc. Class C is rolling out Gemini Tasks in six countriesthe United States, Japan, South Korea, Canada, Australia, and Indiawith initial coverage expected to reach approximately 250 million users. Unlike traditional AI chatbots, Tasks can not only understand user instructions but also connect to external applications to execute specific actions, including calling merchants, waiting for a human representative to answer, booking services, comparing product prices, and assisting with shopping and payments. For example, users can ask Gemini to contact a merchant to book a service and select a suitable time based on their Google Calendar schedule; they can also have it compare different products and complete authorized purchases through Google Pay. On the payment side, Gemini Tasks supports setting spending limits for shopping and processes transactions on supported bank cards through payment tokenization technology. Alphabet Inc. Class C also plans to integrate Unified Commerce Protocol (UCP) checkout functionality to further enhance the agent shopping experience. In addition, Gemini Tasks can access relevant account information through a password manager after authorization and execute tasks in a separate, isolated virtual machine environment to strengthen the security protection of accounts and sensitive information. Users can also issue task instructions via text message, images, or email rather than always operating within the Gemini chat interface. Morgan Stanley noted that Gemini Tasks is currently offered free of charge, and users simply need to switch to "Task" mode within the Gemini app or interface without downloading a separate app. This design is expected to significantly lower the barrier for users to try the new feature. 1 billion users versus 12 million users: Alphabet Inc. Class C has a clear promotional advantage Morgan Stanley believes that one of Alphabet Inc. Class C's greatest competitive advantages is the massive user base Gemini has already built. Currently, Gemini has more than 1 billion monthly active users, while according to third-party data, Meta's standalone AI agent app Muse has approximately 12 million monthly active users. This means Alphabet Inc. Class C does not need to rely entirely on users downloading a new app to build an AI agent user base but can directly promote the Tasks feature to existing Gemini users. Although the first batch of Tasks is expected to cover approximately 250 million users, Gemini's overall monthly active user base of more than 1 billion means there is still room to further expand the service scope in the future. Morgan Stanley believes this user base advantage could enable Alphabet Inc. Class C to expand faster than Meta in the agent market. At the same time, another advantage for Alphabet Inc. Class C comes from its own app ecosystem. Gemini Tasks can be deeply integrated with services such as Google Wallet, Google Photos, YouTube, Gmail, Google Calendar, Google Drive, and Google Flights. For example, users can have Gemini extract flight or event information from emails and automatically add calendar entries; they can also search for travel options, compare hotel prices, or use images in Google Photos to identify products they need to purchase. Morgan Stanley believes these unique app-connection capabilities are expected to enable Gemini to gradually form a user experience that differentiates it from other AI agents. However, the bank also emphasized that Meta likewise possesses important competitive advantages. Facebook, Instagram, and WhatsApp have accumulated a large amount of social relationship and user behavior data, which is expected to help Muse provide more personalized services. In the future, whether Meta can further integrate the data and functions of its social platforms will become an important factor determining Muse's competitiveness. Gemini has connected more than 60 apps; competition in the agent ecosystem is heating up In addition to its own apps, Alphabet Inc. Class C is also accelerating the integration of Gemini with third-party services. Morgan Stanley noted that Gemini Tasks currently has more than 60 app connections, covering areas such as shopping, dining, travel, ticketing, communications, and entertainment. Among them, Instacart can help users add groceries to their shopping cart, including identifying products from images; OpenTable can help users find restaurants and complete reservations; Ticketmaster can be used to search for events and purchase tickets. In addition, Gemini can connect with apps such as WhatsApp and Spotify to help users handle messages, find music, and create playlists. However, some operations still require redirecting to partner websites or apps to complete, meaning not all transactions can be executed directly within Gemini. By comparison, Meta Muse currently has more than 100 app connections, temporarily leading Gemini in number. Morgan Stanley believes that what will determine the competitive landscape of AI agents in the future is not just the number of connected apps but the breadth and depth of integration and the stability of the user experience. The bank noted that the AI agent market may form a mutually reinforcing growth model: platforms with more users are more likely to attract merchants to integrate, and integrating more merchants can further increase user frequency and retention. In this process, Alphabet Inc. Class C's existing massive user base may help it build a commercial ecosystem faster. Gemini's new model improves efficiency by approximately 4x; AI service costs are expected to decline Morgan Stanley believes that, in addition to user scale and app ecosystem, the operating efficiency of AI models will also become a core factor determining the commercialization prospects of agents. The bank estimates that Alphabet Inc. Class C's newly launched Gemini 4 Hydrogen model has approximately 4.28x higher overall efficiency compared with Gemini 3.8 Flash released in September this year. Specifically, the per-task cost of Gemini 3.8 Flash is approximately $7.77, while Gemini 4 Hydrogen reduces it to $3.16, an improvement in cost efficiency of approximately 2.46x. The new model's score on the Artificial Analysis Intelligence Index rose from 31 to 54, further improving intelligence performance relative to cost. Combining the improvement in cost and model capability, Morgan Stanley derived an overall efficiency increase of approximately 4.28x. The bank noted that AI agents need to continuously call models, connect to apps, and execute complex tasks, so computing costs may be significantly higher than traditional chat services. Therefore, whether unit task costs can be reduced while improving model capabilities will directly affect whether AI agents can be promoted at scale and whether related businesses can achieve attractive investment returns. It is worth noting that Meta previously also stated it plans to improve the cost efficiency of Muse's underlying model by 10x by 2027. Morgan Stanley believes this means competition between Alphabet Inc. Class C and Meta will not only be reflected in product features and user scale but will also extend to model efficiency and service costs. $30 trillion consumer market yet to be developed, but Alphabet Inc. Class C faces the risk of search business cannibalization Morgan Stanley believes the long-term commercial value of AI agents may far exceed the current chatbot market. The bank estimates that more than $30 trillion of global offline consumer spending presents an opportunity for further digitization through AI agent services. As agents gradually gain the ability to autonomously search, compare, book, and purchase goods, consumers may increasingly complete transactions through AI assistants rather than visiting multiple websites or apps separately. This change is expected to create new commercialization opportunities for AI platforms and become an important source of returns on tech giants' massive generative AI investments. However, for Alphabet Inc. Class C, the commercialization of AI agents also presents a special challenge: how to create new transaction revenue while avoiding disruption to its existing search advertising business. Morgan Stanley estimates that Alphabet Inc. Class C's search business currently achieves an effective monetization rate of approximately 20% to 40% of relevant transaction value, while the transaction commission rates currently discussed by Meta Muse and other AI agent platforms are only in the low single-digit percentages. This means that if users shift from traditional search to AI agent shopping, Alphabet Inc. Class C may face the risk of growing transaction volume but declining monetization per transaction. Morgan Stanley believes Alphabet Inc. Class C needs to prove that Gemini Tasks can bring additional user engagement and consumption behavior, rather than merely shifting transactions originally completed through search to the AI agent platform. If the new product causes significant erosion of traditional search advertising revenue, investors may reassess the long-term growth capability of Alphabet Inc. Class C's core business and lower its valuation. Conversely, if Gemini Tasks can create incremental transaction demand while maintaining the growth resilience of the search business, it would help increase market confidence in Alphabet Inc. Class C's long-term profitability. Morgan Stanley maintains $375 price target; AI agent competition expected to further intensify in 2027 Morgan Stanley believes the launch of Gemini Tasks marks the consumer AI agent market gradually shifting from technology demonstrations to practical commercial applications. Looking ahead to 2027, tech giants such as Alphabet Inc. Class C, Meta, and Amazon.com, Inc. are expected to engage in even fiercer competition around user scale, app connections, personalized services, and transaction monetization. Among them, Alphabet Inc. Class C's advantages lie in its massive existing user base, comprehensive app ecosystem, and continuously improving AI model efficiency; Meta may achieve differentiation through social data and personalization capabilities. However, Morgan Stanley emphasized that consumer AI agents are still in the early stages of development, and the number of app connections, service quality, and commercialization models all need continuous improvement. Regarding Alphabet's investment outlook, the bank maintains an "Overweight" rating and a $375 price target. This price target is based on the bank's forecasts of Alphabet's earnings per share of approximately $15 and $18 for 2027 and 2028, respectively, and applies a P/E ratio of approximately 23x to the two-year average expected earnings per share. In a bull case, if AI innovation drives accelerated growth in the search advertising business, Alphabet Inc. Class C's cloud business performs further improvement, and new AI products bring additional revenue without significantly eroding the search business, Morgan Stanley believes Alphabet's stock price could reach $450. In a bear case, if global advertising growth slows, AI product monetization falls short of expectations, and higher computing costs pressure margins, its stock price could fall to $225.