ChatGPT is being used to disrupt elections around the world ...
Hacking groups affiliated with regimes in China, Iran, and Russia have been named as suspects in a 54-page report. Despite ongoing polls showing the election as a dead heat, there are concerns about potential interference through the misuse of artificial intelligence tools like ChatGPT and Dall-E.
OpenAI has uncovered 20 campaigns worldwide using its AI tools, ChatGPT and Dall-E, to influence elections. These manipulative activities range from generating fake personas and posting content on social media to crafting articles for websites. The emergence of generative artificial intelligence has raised alarms about how it could be used to sway elections, prompting leading firms to take preventive steps.
Threats to Election Integrity
In a testimony before Congress, OpenAI's CEO, Sam Altman, expressed concerns about the potential threat such generative AI poses to election integrity. This year, in particular, highlights the need for robust defenses against cyber actors and covert influence operations that aim to leverage AI models for deceptive campaigns on various online platforms.
According to OpenAI's latest report, they have already disrupted more than 20 operations and deceptive networks globally that attempted to utilize their models for malicious purposes. Hacking groups linked to regimes in China, Iran, and Russia have been singled out as potential suspects in these interference campaigns.
Case Studies and Operations
The report delves into specific case studies, including an operation attributed to a "Russia-origin threat actor" that targeted West Africa and the UK. This operation utilized AI models to produce content in English and French, which was then disseminated through websites masquerading as news outlets in the two regions.
The UK-focused "news" brands established partnerships with local organizations, such as a church in Yorkshire, a school in Wales, and a chambers of commerce association in California. These efforts aimed to build an audience and disseminate misinformation through various channels.
While the report does not disclose the specific organizations involved, OpenAI continues to monitor and disrupt such deceptive operations to safeguard election processes from undue influence.
Join thought-provoking conversations, follow other Independent readers, and see their replies. Please refresh the page or navigate to another page on the site to be automatically logged in. Alternatively, refresh your browser to be logged in.
