Tech News

Foreign propagandists continue to use chatgpt in impact campaigns

China’s publicity and social engineering operations to create posts, comment and promote engagements at home and abroad. Openai said it recently undermined the secret influence of four Chinas, operations that are using its tools to generate social media posts and responded to platforms such as Tiktok, Facebook, Reddit and X.

These comments revolve around several themes from American politics to Taiwanese video games, where players fight against the Chinese Communist Party. Chatgpt is used to create social media posts that both support and condemn different hot button issues to inspire misleading political discourse.

Ben Nimmo, principal investigator at OpenAI “What we see from China is the use of increasingly more tactics and increasing number of covert operations.” While Openai claims it also undermines a few actions it believes originated in Russia, Iran and North Korea, Nemo detailed the actions of China, which they “target many different countries and themes.” […] Some of them combine elements of influence, social engineering, surveillance. ”

This is far from the first time. In 2023, it was discovered that since 2019, AI-generated content has been used in politically motivated online influence campaigns.

In 2024, efforts to undermine five state-affiliated operations of China, Iran and North Korea are outlined, which are using OpenAI models to achieve malicious intent. These applications include debugging code, generating scripts, and creating content for phishing activities.

That same year, Openai said it was through Chatgpt that created long-term political articles about the U.S. elections, and then posted conservative and progressive channels on fake news sites. The operation also posted comments on X and Instagram via fake accounts, again embracing the opposite view.

“We don’t usually see these operations getting more and more due to using AI,” Nimmo . “For these operations, better tools don’t necessarily mean better results.”

It’s hardly comfortable. With generative AI being cheaper, and reasonably thought its ability to generate content together would make the impact of similar activities easier and more affordable, even if their efficacy remains the same.

Related Articles

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button