Microsoft hits back at claims AI data scraping was sneakily turned on in Word, Excel
Recent controversies questioned whether Microsoft harvested private customer data from Office applications to train its AI models. Critics argued that default settings for connected experiences potentially allowed for unauthorized scraping of proprietary content, raising significant concerns regarding creator rights and data privacy in the age of generative AI. Microsoft has firmly denied these allegations, clarifying that customer data is never used to train large language models. Instead, the company emphasized that these connectivity features solely facilitate functional aspects like real-time collaboration. They further assured users that proprietary information remains secure and private, distinct from public data sources used for model development. This distinction is crucial for open data discourse, as it highlights the ongoing tension between utilizing freely available public information for AI training and protecting sensitive, user-generated private data. It underscores the importance of transparent data policies and clear opt-out mechanisms in maintaining trust between technology providers and their users.
Source: techradar.comPublished on 2024-11-28
Related news
- Barings Law plans to sue Microsoft and Google over AI training data | Computer Weekly
- Bluesky Open API: Data Scrapers May Access Firehouse for AI Training, as Demoed by Hugging Face
- Abu Dhabi’s Technology Innovation Institute Inaugurates Open-Source AI Summit with Critical Discussions on the Future of AI
- One million public Bluesky posts scraped for AI training
- La transparencia no solo revela, también transforma; llama IACIP a defender el acceso a la información