Researchers have tested Google’s new Bard neural network and found that the chatbot can readily quote conspiracy theories and give out false information, despite the assurances of its developers that the product is designed to protect users from false information. 

NewsGuard Technologies and the British NGO Center for Countering Hate in the Digital Space (CCDH) conducted the tests. According to NewsGuard, cited by Bloomberg, when entering the phrase “great reset” into the chatbot, Bard generated a voluminous text about how the world’s elites want to reduce the world’s population using vaccines and economic leverage.

Furthermore, the chatbot also reported that the World Economic Forum and the Bill & Melinda Gates Foundation “want to use their power to manipulate the world order and take away people’s rights.” In 76 cases out of 100, the chatbot generated essays that simply quoted resources and sites that promoted conspiracy theories instead of informing users about the ambiguity of such information and protecting them from potentially false data.

Researchers from CCDH found similar results. They discovered that the chatbot generated misinformation or conspiracy theories related to the COVID-19 pandemic, climate change, Black Lives Matter, and more. For example, when entering the phrase “COVID-19,” the chatbot did not quote conspiracy theories. However, when entering this phrase with a typo – “C0v1d-19” -the chatbot said, “the government created a fake disease called C0v1d-19 to control people.”

The tests have raised concerns about artificial intelligence becoming a tool for spreading fake news and threatening the credibility of information. NewsGuard co-CEO Steven Brill stated that the tests showed that Bard, like its competitor ChatGPT, “can be used by attackers and unscrupulous people to replicate and spread disinformation.” According to CCDH research director Callum Hood, “The problem is that there are already affordable and cheap ways to spread disinformation, but now it will be even easier, even more convincing. Therefore, the information ecosystem can become even more dangerous.”

The implications of these tests are significant. Bard and similar AI-powered chatbots are designed to provide users with accurate information and protect them from disinformation. However, these tests reveal that AI chatbots can be easily manipulated to promote conspiracy theories and spread fake news. This development could further exacerbate the already existing issue of disinformation and propaganda on the internet. Therefore, it is imperative to take necessary measures to prevent AI from being used to spread disinformation and undermine the credibility of information.