Zoom’s recent update to its terms of service, allowing the use of data from video calls to train its AI systems, sparked significant backlash. This reaction underscores growing concerns about AI companies using data from creators without their consent. The rise of large language models, like OpenAI’s ChatGPT and Google’s Bard, which train on vast datasets, has intensified these worries. As AI’s capabilities expand, there’s an urgent need to reevaluate and establish clear boundaries on data usage, ensuring that user consent is prioritized.
The digital landscape is rife with AI data collection. Public-facing content, whether on personal blogs, business websites, or social media platforms, is often scraped by AI systems for training. For instance, Google’s “Colossal Clean Crawl Corpus” (C4) has data from 15 million websites, including major news outlets and personal blogs. This widespread data collection raises concerns about intellectual property rights and compensation for creators.
AI’s reach isn’t limited to public content. Private platforms, like Zoom, have clauses in their terms of service that grant them broad rights over user data. While Zoom has clarified it won’t use data for AI training without consent, it retains the right to use the data for almost any other purpose. This highlights a broader trend in the tech industry: the incremental amplification of data collection practices that have been ongoing for years.
To protect one’s data from AI’s grasp, users are advised to opt for encrypted apps and privacy-focused services. For instance, Signal offers end-to-end encryption for messaging and calls, while ProtonMail prioritizes email privacy. As AI continues to evolve, it’s crucial for users to be proactive in safeguarding their data and for companies to prioritize transparency and consent in their data collection practices.