At the Google Cloud Next conference held on Wednesday, Google announced that it will introduce the intelligent proxy function with "Auto Browse" capability into the enterprise version of Chrome browser, and simultaneously strengthen relevant security protection measures. With this feature, enterprise users can use Gemini to understand the real-time web content in the currently opened tab, and then let AI perform various tasks related to web operations such as booking travel, entering data, arranging meetings, etc.
Google said that the automatic browsing function can help employees complete a variety of typical scenarios: for example, based on the content in Google documents, enter key information into the company's designated CRM system; automatically compare quotes from different suppliers in multiple tabs; quickly summarize the candidate's portfolio before the interview; extract key data from competitors' product pages, etc. However, these workflows still require "humans in the loop", that is, the user must conduct manual review and confirmation before the AI generates input or action, and the final decision-making power remains in the user's hands.
Google emphasizes that the goal of these capabilities is to speed up tedious, mechanical transactional operations, thereby freeing up employee time so that they can devote more energy to what the company calls more "strategic" work. This logic is also in line with the grand promise of current AI supporters: to "help you take back time" through new technologies. However, actual research shows that AI does not really reduce workload in many scenarios, but has a tendency to "increase work intensity." As for how AI will evolve when it becomes part of standardized workflows in an enterprise-level environment, it remains to be seen. Predictably, managers may expect employees to complete more tasks in the same amount of time.
Google said that the new feature will be first available to Workspace users in the United States as part of its strategy to deeply embed AI into "standard applications" in the workplace - the web browsers used by almost everyone. Enterprise administrators can enable this feature through policy configuration. Google also promised that the prompt information entered by users within the organization when using these functions will not be used to train its AI models. At a time when the outside world is becoming increasingly sensitive to the use of data, such disclosures have become necessary, especially in the context of Meta's revelation that even employee keyboard input is used to train AI models.
Similar to the consumer version, Enterprise Chrome users can save frequently used workflows for quick recall later. These preset workflows are called "Skills" by Google. Users can either call up the Skill to be used by inputting a slash "/", or click the plus button on the interface to select it, thereby quickly calling up common processes in complex web page operations.
While enhancing Chrome's AI capabilities, Google is also strengthening its detection and control of "unauthorized AI tools." With Chrome Enterprise Premium, Google has previously been able to identify unauthorized AI tools used within enterprises. Now this ability will be extended to monitor compromised browser extensions or other AI services, focusing on "anomalous agent activity." From a security perspective, this is positioned as helping IT teams identify potential risks, but it also has another meaning: Google is leveraging the power of enterprise IT to curb other AI agent services that have grown naturally in the workplace from the bottom up.
In the past "Enterprise 2.0" wave, many cloud storage services, collaborative documents or file sharing tools gradually gained a foothold within the enterprise through voluntary adoption by employees. Now, Google is trying to control the entry and use scope of generative AI and various SaaS in enterprises through its official tool suite and monitoring capabilities. Google named this new capability "Shadow IT risk detection", which allows IT teams to fully understand the actual use of authorized and unauthorized generative AI and SaaS websites within the organization.
At the same time, IT teams will also receive a Gemini Summary provided by Gemini, which provides an overview of the Chrome Enterprise release notes and other change information. This summary highlights key changes, new policies, and retiring features, along with AI recommendations for configuring new settings, auditing managed browsers, and more.
In the field of security, Google also announced an expansion of cooperation with Okta to strengthen the security protection of the "agent workplace" and add more mechanisms to reduce the risk of attacks such as session hijacking. In addition, Google has also upgraded its security management capabilities for browser extensions and introduced integration with Microsoft Information Protection (MIP) to help enterprises implement consistent security policies across multiple systems.