Meta Says Generative AI Can Reduce Negative Content During The World Election Period

Nick Clegg, President, Global Meta Affairs at the World Economic Forum (photo: Meta)

JAKARTA - Currently, artificial intelligence (AI) and AI technology are generating, being swords and shields for the entire world community, where this generating AI not only brings benefits but also concerns.

In a panel discussion at the World Economic Forum held in Davos, Switzerland, Nick Clegg, President, Global Meta Affairs expressed concerns about the potential abuse of generative AI equipment during the election period.

"We have discussed with experts about the impact of AI's progress ahead of this year's election, and we have policies that we can implement regardless of whether content is created by AI or humans," Clegg said.

While currently not many countries have regulations that set formal industry standards in AI use and development, Meta is taking independent steps to keep its platform safe.

One of them is, helping people understand when the image was made with the AI Meta feature, using watermarking.

I can easily imagine in the next few years, we're just talking about watermarking on synthetic content. I can easily imagine one day we'll all be looking for watermarking for original and verified content," Clegg said further.

According to him, the watermarkingi tool can provide certainty on the internet, so that people who see the content can prove the authenticity of the content.

On Facebook, Clegg mentioned that their generative AI tools are able to significantly reduce AI-made content or hybrid content.

"The only thing I want to say about AI, yes of course generative AI can accelerate the application of synthetic materials and so on. It's also one of the best weapons, on distribution platforms like Meta, to really identify bad content," he explained.

So if you look, for example, the prevalence of hate speech on Facebook right now, explains Clegg, you may only find 0.01 percent.

The number is reduced, so every 10,000 bits of content, you may see a little hate speech. This decline is very significant in the last two years for one reason alone, one reason only, namely AI," he said.

"AI has become a very effective tool, an improvement in classification in pursuing bad things. So it's swords and shields. I think we need to remember that," he concluded.