Zuckerberg wants to be relevant in the Trump administration

On the issue of freedom of expression, Clegg acknowledged that it is complex to find the balance between allowing people to make their voices heard and maintaining their safety. “It’s something that no platform will do 100 percent right all the time,” he said.

Among the most relevant actions in this regard are allowing people to ask questions or express their concerns about the electoral processes in organic content.

However, there was no room for claims or speculation about corruption, irregularities or bias in the elections when combined with signals indicating that the content was intimidating violence. Likewise, with respect to paid content, advertisements that questioned the legitimacy of an election were prohibited.

Additionally, annual audits were conducted on words that the company designates as profanity under its Hate Language policies, with the goal of maintaining updated lists that protect vulnerable communities and ensuring no mistakes are made in moderating public speech. .

The impact of AI on elections

With the rise of generative Artificial Intelligence, concern about its impact on elections also increased and according to Meta there were cases of confirmed or suspected use of AI.

However, volumes remained low. According to data from the Meta AI platform, 590,000 requests to generate deepfakes of President-elect Trump, Vice President-elect Vance, Vice President Harris, Minnesota Governor Tim Walz and President Biden were rejected in the month leading up to Memorial Day. the elections.

“Our existing policies and processes have proven sufficient to reduce the risk around generative AI content. During the election period, ratings on AI content related to elections, politics and social issues represented less than 1% of all verified disinformation,” said Clegg.

The company reported that on monitoring the use of generative AI by covert influence campaigns (also known as coordinated inauthentic behavior networks or CIBs), which were found to have achieved incremental improvements in productivity and the generation of content.

To combat this issue, the company said it has focused on dismantling influence operations, through the elimination of campaigns and not on the content they publish, whether created with AI or not.

While Clegg celebrated the progress they made this year, he also warned that the prevalence of these AI tools “will clearly continue to grow in scale over time” from a disinformation point of view, but also useful and even playful.