Trump's anti-woke ordinance: Meta works with right-wing influencer
To get the supposedly woke bias out of AI applications, Meta is now working with the conservative Robby Starbuck.
(Image: Michael Vi/Shutterstock.com)
AI is too woke, says Donald Trump. That is why the US President issued an executive order requiring every AI company to make its AI applications less politically correct. Meta is now complying with this. Robby Starbuck becomes new advisor. The influencer is highly conservative and had just tried to sue Meta.
Meta AI, Meta's AI chatbot, allegedly said that Starbuck was present at the violent protests at the Capitol on January 6. According to the influencer, this is not true, which is why he is suing. He is also referring to Donald Trump's anti-woke ordinance, which Meta is said not to have complied with. According to the Wall Street Journal, part of an agreement between the parties is now that Meta will hire Starbuck. In a statement to the newspaper, Meta and Starbuck said: “Since Meta engaged with Robby on these important issues, the company has made tremendous progress in improving the accuracy of Meta AI and reducing ideological and political bias.”
It is unclear whether Meta is also paying Starbuck for his work or as part of the settlement. Just a few weeks ago, Meta paid Donald Trump 25 million US dollars to settle a legal dispute over the blocking of Trump's account after the storming of the Capitol.
Videos by heise
Starbuck makes YouTube videos in which he criticizes companies such as Toyota, John Deere, and Harley Davidson for having so-called DEI departments. These are teams that take care of diversity, equity, and inclusion. His followers should avoid these brands.
Data protection versus AI models
The legal handling of false information about a person provided by an AI chatbot has not yet been conclusively clarified. There has already been a court case in the USA in which a radio presenter sued OpenAI. The judge ruled in favor of OpenAI. In the EU, the GDPR actually stipulates that people have the right to have false information about them deleted. However, this is not possible in this form with an AI model. False statements can neither necessarily be reproduced nor definitively prevented.
(emw)
