Meta: AI content was less than 1% of election misinformation

After people warned about the potential impact of generative AI on the US Election, Meta released a study showing AI-generated content represented less than 1% of all fact-checked misinformation.

In a blog post, the technology giant states they were looking across their services and monitoring the impact of AI throughout this time.

Their findings suggest the risks “did not materialize in a significant way and that any such impact was modest and limited in scope.”

There were instances of confirmed or suspected use of AI, but the volumes remained low and the company states its “existing policies and processes proved sufficient to reduce the risk around generative AI content.”

Meta also prevented people from using its Imagine AI image generator to create election-related deepfakes. It rejected 590,000 requests to generate images of President-elect Trump, Vice President-elect Vance, Vice President Harris, Governor Walz, and President Biden in the month leading up to election day.

It re-adjusted its AI chatbot too, so it would respond to questions about when and how to vote with links to authoritative sources that could offer the most up to date information. On the actual Election Day, the chatbot redirected people to check with sources when asked questions about the candidates and results.

With social media now having a stronghold over daily life, the tech company which owns Facebook, Instagram, Threads, and WhatsApp says: “Striking the balance between allowing people to make their voices heard and keeping people safe is one that no platform will ever get right 100 per cent of the time.”

They state their error rates “are too high, which gets in the way of the free expression we set out to enable.”

Those in the US have the choice of whether they want to see more political content recommended to them or otherwise, as political content controls were rolled out on Facebook, Instagram and Threads.

The company is in the process of rolling this feature out globally.

Image Credit: Via Meta Blog

The post Meta: AI content was less than 1% of election misinformation appeared first on ReadWrite.