Watermarking seems a basic solution, but if companies that help produce AI content like ChatGPT or Midjourney, or companies that distribute them like Meta or X can collaborate towards watermarked content, it will be much harder for AI to confuse or misdirect the public.

And that process is understandably hugely important. So much so that tech giant Intel is leading the charge with their new tool – FakeCatcher. It’s a big step in the right direction, but it won’t stop there. There’s a huge need to start watermarking AI content, helping us easily tell human-made from AI-made, and giving everyone a clear picture of what’s what.

Whilst we can individually watermark our own content, like in this article, ultimately to enforce it en masse, it has to be done by the companies that produce and share the content. Doing would be like a big neon sign saying, “We’re committed to the truth!” – reducing the potential harms of AI content.

‍

Realistically we have an incredible, blossoming technology at our fingertips.

The ability to summon previously inaccessible information and more from the ether still feels like magic- and that’s because the capabilities are just so enthralling.

But like when the internet rolled out across dialup connections around the world, or social media hit every young persons’ palm, there’s a huge ethical weight to the exhilarating possibilities being mapped.

We need everyone to contribute to safe digital communities. Nothing has changed there, but the tools are different and certainly more powerful than ever.

With the right framework it will be easy. And better still, you’ll see your day-to-day tasks drop away from the drudgery and into the more intriguing, the novel and the curious. It’s exciting and unfolding right before your eyes.