Microsoft speaks plainly: users should be made “addicted” to AI

Microsoft has introduced Scout, a new AI assistant for office applications. Leaked internal documents now show the exact strategy behind the software: The main goal of the first phase is to get users “addicted”.

Internal strategy for Scout

Microsoft presented some new products at the Build developer conference this week. As expected, the focus was of course on AI. Project Solara is working on, among other things, the AI ​​hardware of the future and bringing OpenClaw to Windows. Under the name “Scout”, the open source project will be deeply integrated into Microsoft 365 and will relieve users of everyday tasks. The program operates independently on the desktop, manages the calendar or answers emails.

Scout marks the next big step for the group in the automation of typical office work. However, leaked internal papers now reveal the company’s exact introduction strategy for the new tool. Microsoft doesn’t mince words and describes a three-stage plan in the documents, the first phase of which, entitled “Get people addicted,” defines a very clear goal: The software should be explicitly designed in such a way that users become addicted. Scout: Microsoft introduces its new personal AI assistant

How 404 Media reports that their own employees view this choice of words as worrying. An anonymous Microsoft employee explained that dependency should never be part of a development strategy. One software developer noted that creating addictive programs is the end goal of many large technology companies. Such mechanisms are known from social networks, where algorithms are intended to keep users on the platform for as long as possible through constant rewards. Microsoft is now obviously making this principle its guiding principle at Scout.

Test run with the CEO

The development of Scout is led by Omar Shahine. Microsoft’s corporate vice president originally programmed his own assistant in his free time and presented the concept in an internal Microsoft funding program. The management was convinced of the potential. He was then given the task of working with a team to turn the prototype into a fully-fledged product for the mass market. The company has been intensively testing the tool internally since March. Currently, over 1,000 employees use the software every day, including CEO Satya Nadella. According to the papers, the internal tests show a very long service life and strong retention rates. The assistant therefore spread completely organically and without formal marketing to become the most desired internal application within the workforce.

Safety is secondary for now

In order to be able to complete tasks such as filling out expense reports or booking flights, artificial intelligence requires extensive access rights to sensitive accounts. This represents a significant risk to data protection. However, the leaked papers only treat the topics of security and compliance with company guidelines as minor points that the development team will still have to clarify in the future. Especially in view of the growing criticism of the power of chatbots, this prioritization carries risks. While many competitors pursue similar user retention goals, explicitly codifying a dependency strategy in writing is a delicate step. However, it once again sheds light on how modern tech companies view their customers.