How Anthropic's Watermarking Plan Falls Short of Transparency The European Union's AI Act has been a thorn in the side of the tech industry, forcing companies to acknowledge and label their AI generated content.
Anthropic's recent announcement that it will add watermarks to its AI models, including Claude, is just the latest example of this trend.
At first glance, watermarking seems like a straightforward solution to a complex problem. However, upon closer inspection, it reveals a more insidious dynamic at play.