These days, I notice that my Instagram feed has become flooded with reels on how to get around a watermark on content that has involved AI in some way. Anthropic has deployed this invisible watermark as part of complying with emerging transparency requirements, including the EU AI Act, and has extended the feature across its platforms.
And that has users quite infuriated.
Watermarks are not new. Google has used its own SynthID, an invisible digital watermarking technology developed by DeepMind. It embeds imperceptible markers directly into the pixels of images and videos, the frequencies of audio, or the token-generation patterns of text created by AI models like Gemini. This was rolled out over 2023 and 2024.
Then along came Anthropic, using a version of Google DeepMind’s SynthID-Text approach. But somehow, the timing, wide implementation and the fact that it was announced amid heated anxiety about AI detection and professional exposure had users up in arms.
On top of everything else, Anthropic acknowledged that putting your own work into, say, Claude for substantial editing could leave enough of a watermark to be detected. The watermark itself can’t distinguish between content Claude generated and something written by the user that Claude heavily edited. The watermark comes from a genuine intention to be transparent, but it bumps into the untidy reality of how people actually use AI for content.
Writers’ favourite
Claude has built a reputation as a premier AI tool for natural, nuanced writing, long-form editing and complex coding. Many users deliberately switched from other AIs to use Claude. The watermark directly affects those for whom writing and coding are professions, raising questions of authorship, attribution and professional trust. Pro and enterprise users paying significant monthly amounts now face the possibility that substantially edited work may carry a watermark when it goes to a client. Obviously, it all depends on what you use AI for.
The absurdity of this situation is that we keep inventing technology that is breathtakingly smart, but let it loose into the wild without thinking through the implications. Much energy has been spent on fine-tuning and improving how AI tools write, showcasing their skills and positioning AI as the ultimate ‘collaborator’ and colleague, as the new brain that you can use to enhance your own work. Now that users begin to do exactly that, their work is treated with suspicion.
Writing co-created by a user and an AI tool like Claude may be based on impressive contributions from both. The idea may come from AI and its brilliant execution from the user. Or the idea may be quite original and AI may help shape it or nudge it to completion in the right direction. It could help make the best of a great idea. But that raises the question of who should receive credit for the result.
Blurred lines
Meanwhile, there are two other issues muddying the waters. The first is that some users do get an AI chatbot to write substantive work for them and only contribute their name to it. Claiming authorship of work one didn’t create is not just unethical, but it’s also the kind of outsourcing that could erode thinking skills.
The other issue is that using AI to churn out low-quality, quick content is swamping the internet with slop. LinkedIn found it has a problem with AI slop, with users getting fed up with low-effort content. It’s AI-generated material that sounds polished but contains little original perspective, experience or substance. The platform therefore added a “Seems like AI slop” option. More than a million people used it within the first few weeks.
But importantly, clicking it doesn’t automatically punish the author. LinkedIn combines that feedback with other signals in its detection and ranking systems. LinkedIn says suspected AI-slop content is now getting about 40% less visibility than it did a few weeks earlier.
AI-generated content isn’t necessarily slop, and all slop isn’t necessarily AI-generated. But perhaps this tension has come up at a time when both AI and its users can make course corrections to how they work together—to crowd out slop and replace it with content of substance.
In the process, they can make the internet great again.
The New Normal: The world is at an inflexion point. Artificial intelligence (AI) is set to be as massive a revolution as the Internet has been. The option to just stay away from AI will not be available to most people, as all the tech we use takes the AI route. This column series introduces AI to the non-techie in an easy and relatable way, aiming to demystify and help a user to actually put the technology to good use in everyday life.
Mala Bhargava is most often described as a ‘veteran’ writer who has contributed to several publications in India since 1995. Her domain is personal tech, and she writes to simplify and demystify technology for a non-techie audience.