Instagram app on an Android phone.
Credit: Reuters File Photo
Deepfakes have been around for close to a decade. Back then, it was treated as a fun, innovative way to create fun content. But, with growing cases of misuse to defame people and also steal money from naive smartphone users, there is an urgent need for regulations.
What's more worrying is that with the advancement of generative Artificial Intelligence technology, it has become increasingly hard to detect real from deepfakes.
Now, Instagram Chief Adam Mosseri, in a new year message on the Meta-owned platform, has flagged concerns over AI-generated content. While appreciating the creative AI slop content, he warned of the need for watermark-like tags so that it can be differentiated from real.
For the uninitiated, 'AI Slop' content is a generic mass-produced synthetic content created for multimedia consumption. In India, 'Bandar Apna Dost' is the most popular AI slop content channel on YouTube. It shows a talking monkey travelling the popular tourist spots with a running commentary on life. It has a massive fan following with more than 27.6 lakh subscribers and billions of minutes of views. However, YouTube doesn't offer a monetisation option for such fully AI-generated content.
That said, the uncontrolled circulation of such content is not advisable.
"Platforms like Instagram will do good work identifying AI content, but they'll get worse at it over time as AI gets better. It will be more practical to fingerprint real media than fake media," said Mosseri.
"Labelling is only part of the solution. We need to surface more context about the accounts sharing content so people can make informed decisions. Who is behind the account?. Also, we need to build the best creative tool. Label AI-generated content and verify authentic content. Surface credibility signals about who's posting. Continue to improve ranking for originality. Instagram is going to have to evolve in a number of ways, fast," concluded Mosseri.
While government regulators are taking their own sweet time, technology companies should form an alliance to bring a standard rule to integrate watermark and metadata with every single piece of synthetic content churning out of their respective apps.
Big firms like Apple, Google, Meta, and OpenAI are already following this process; lesser-known publishers have failed to comply to the standards.
Get the latest news on new launches, gadget reviews, apps, cybersecurity, and more on personal technology only on DH Tech.