CybserSecurity News

Tag: privacy sandbox

  • Google has started disabling third-party cookies for Chrome users

    Google has commenced the process of phasing out third-party cookies in its Chrome browser, a change that, since January 4th, has affected 1% of users worldwide as part of the “Privacy Sandbox” project. Considering Chrome’s market share, this deactivation is estimated to impact around 30 million users.

    Users included in this initial wave of deactivation will receive a notification indicating their participation in an experimental privacy protection initiative. This feature restricts the use of third-party cookies for online activity tracking. Google has provided the option for temporary re-enabling of third-party cookies should users experience functional issues with websites.

    CVE-2022-4135

    Google’s “Privacy Sandbox” initiative is designed to offer an alternative to traditional cookies. This system categorizes users based on their browsing activities into interest groups to deliver targeted advertising, while ensuring that data processing occurs directly on the user’s device. Google intends to retain information about user interests for three weeks.

    This project has drawn regulatory attention due to concerns that it may strengthen Google’s market position. Should the implementation prove successful, Google plans to gradually expand this tracking protection over several months, to eliminate third-party cookies for all Chrome users by mid-2024.

  • Google simplifies the ad creation process with auto-generated artificial intelligence

    In the wake of numerous product and service announcements integrating generative artificial intelligence technologies at Google I/O 2023, Google has affirmed its commitment to further streamline the advertising process through the application of artificial intelligence, enhancing the efficacy of ad exposure.

    Leveraging natural language understanding capabilities, Google can now employ artificial intelligence to analyze and summarize the content of target web pages, automatically generating relevant and effective ad keywords, titles, images, and related descriptions. This fosters the creation of advertising content perfectly tailored to the exposure needs of the target web page.

    Google ads artificial intelligence

    Simultaneously, Google preserves the flexibility for ad managers to manually adjust, edit, or remove unsuitable content, preventing unnecessary impacts.

    As Google Search service begins to incorporate generative artificial intelligence, producing corresponding content in search results automatically, Google also plans to present more pertinent advertising content on the search result display panel based on the user-inputted search keywords.

    Google will enable ad managers to use conversational descriptions in the Google Ads service backend to facilitate the system’s automatic generation of suitable ad material. Further, Google employs generative artificial intelligence to allow ad managers to create more eye-catching ad images through automated image generation technology, even improving the clarity of originally blurry or low-resolution images, or presenting products in a clean cutout style, thus refining the ad content.

    In conjunction with the Smart Bidding feature and the high-efficiency ad campaign feature enhanced by artificial intelligence, Performance Max, Google asserts that under identical ad bid situations, conversions can increase by over 18%, a growth rate over 13% higher than the first year this technology was introduced.

    Google indicates that as the search service integrates generative artificial intelligence technology, advertising services are expected to evolve in tandem. This evolution will make the ad placement process simpler while enhancing the precision and effectiveness of placements. It can even enable laypeople to create more effective ad content through conversational descriptions, allowing more businesses to boost product exposure through the Google ad system.

    Moreover, Google emphasizes increasing transparency in the use of its advertising products, which includes using distinct fonts to separate ad content from natural search results. For instance, the use of the tag “Sponsored” allows users to discern ad content, and users can even adjust the ad content display mode through settings.

    As the Chrome browser will gradually phase out the use of cookies starting next year, and replaced them with the Privacy Sandbox, Google asserts that ad content will continue to be precisely targeted to specific audiences in a more secure manner that does not infringe upon user privacy.

  • Google will adopt a new version of the privacy sandbox from Q1 2024

    Google recently provided further elucidation of its advertising privacy policy, affirming that it will begin implementation in the Chrome browser during the first quarter of 2024. Users will gradually transition to the new operational model of the Privacy Sandbox, in conjunction with the Topics algorithm, to deliver interest-based advertisements without infringing on user privacy.

    Google first expressed in 2020 its intentions to eliminate the advertising technique of recording user browsing preferences through cookies. The initial proposition of the FLoC (Federated Learning of Cohorts) algorithm sparked controversy, causing Google to postpone its adoption until 2023, subsequently proposing to construct a more privacy-conscious advertising model using the Topics algorithm.

    Google stated in its current elucidation that it aims to completely abandon the use of third-party cookies for recording user browsing behavior in the latter half of this year. Instead, it will adopt the new Privacy Sandbox technology along with the Topics algorithm to serve interest-based advertisements without compromising user privacy.

    According to Google, the Topics algorithm will learn potential interests based on a user’s recent browsing content, subsequently serving appropriate advertisements during browsing. Google emphasizes the avoidance of sensitive topics, while allowing users to opt out of the Topics algorithm or adjust their interest preferences under the new advertising privacy policy.

    Furthermore, Google underscores that under the Topics algorithm, the proportion of user privacy data accessed by the system will be significantly reduced. The operational method is touted as being more transparent, allowing users to maintain greater control over how their private data is utilized.

    Google has already started to incorporate the Privacy Sandbox functionality beginning with the Chrome 101 test version, with plans to officially offer it in the Chrome 115 version, ensuring user browsing privacy while balancing the requirements for advertising content delivery.

  • Google Chrome is testing new privacy sandbox features

    Previously, under pressure from the browser industry, Google set out to improve browser privacy issues. At that time, Google began testing privacy sandboxes to reduce the risk of user privacy leakage.

    The privacy sandbox also collects user information but transfers it to advertising networks for use. Compared with Google’s previous collection of data, there are fewer data available to advertising networks.

    On this basis, Google is studying new privacy security policies to improve the privacy sandbox function, the most important of which is not to upload user data to the server.

    Normally, the real-time bidding system adopted by the advertising network needs to upload user data to the server, so the system can bid normally only after matching the data.

    Due to privacy protection and anti-monopoly issues, Google Chrome does not want to share user data with advertising networks.

    So Google is currently testing a new thing called the Federated Learning of Cohorts, which aims to analyze user data locally and then group it.

    Each group can accommodate thousands of users, these users belong to the same interest category, and Google will only share these categories with advertising networks for advertising.

    This means that user data has not actually been uploaded or directly exposed to the advertising network. The advertising network can only see specific categories of users and cannot be targeted.

    At the same time, tests have shown that this model can “see at least 95% of the conversions per dollar spent when compared to cookie-based advertising.”

    According to Google, Google Chrome will start testing the new privacy sandbox in March but will test it with customers on the Google advertising network in the second quarter.

    If all goes well, the new privacy sandbox will be widely launched in the future, which can greatly reduce the ad network’s acquisition of user data and continuous tracking.

    Of course, this is only for advertising networks. As for whether Google and Google Chrome itself will continue to collect data and store it on Google servers, it is still unknown.

    In addition, Google is also testing a new feature called Trust Token, which aims to store the collected data on a trusted server and output it through a trusted token.

    Ad networks can only read trusted tokens and cannot track users continuously, which will help users avoid ad networks to track them in various ways.