Rédaction Africa Links 24 with Guardian Nigeria
Published on 2024-02-16 05:04:05
Meta, the parent company of Facebook, Instagram, Threads, and WhatsApp, is taking proactive steps to detect and label AI-generated images across its social media platforms. This move comes as the company aims to tackle the issue of deceptive content and promote transparency for its 36.2 million advertising audiences in Nigeria.
Nick Clegg, President of global affairs at Meta, announced in a blog post that the company will begin labeling AI-generated images on its platforms “in the coming months.” This initiative is part of Meta’s commitment to working with industry partners to develop common detection standards for AI content.
While Meta already applies “Imagined with AI” labels for images created using its own Meta AI feature, the company now aims to expand this labeling to include content created using AI tools from other companies. The decision to label AI-generated images is driven by the need to differentiate between human-created and AI-generated content, as well as to provide users with transparency about the use of AI technology.
Clegg emphasized the importance of establishing a clear boundary between AI and human-generated content. He stated that as the line between human and synthetic content becomes increasingly blurred, it is crucial to inform users when they are viewing photorealistic images created using AI. To achieve this, Meta will be developing the capability to detect AI signals and label AI-generated images posted by users across its platforms.
In addition to visible markers, Meta is also exploring the use of invisible watermarks and embedded metadata within image files as measures to identify and label AI-generated content. The company aims to deploy these labeling capabilities across all languages supported by each app, with the goal of enhancing transparency and understanding for its user base.
Clegg highlighted Meta’s collaboration with industry partners to align on common technical standards for identifying AI-generated content. By working with partners such as Google, OpenAI, Microsoft, Adobe, Midjourney, and Shutterstock, Meta aims to implement industry-leading tools that can detect invisible markers at scale and label images created using AI tools from these companies.
The company’s efforts to label AI-generated content coincide with the approach of several important elections taking place worldwide. Meta plans to use this period to gather insights into how people create and share AI content, as well as to evaluate the most valuable forms of transparency for users. This learning will inform the company’s approach to industry best practices and technology evolution going forward.
By taking these proactive steps to detect and label AI-generated images, Meta is demonstrating its commitment to promoting transparency and safeguarding its platforms against deceptive content. The company’s collaboration with industry partners and investment in labeling technology signal its dedication to addressing the evolving landscape of AI-generated content on social media.
