Artificial intelligences have come to turn the world upside down. In a matter of months, the number of users of chatbots like ChatGPT, Bard, or the new Bing has continued to grow, and more and more companies are implementing AI into their services.
This is the case, for example, with Snapchat, Instagram, and TikTok, three of the most widely used social media platforms worldwide, which have been developing their own chatbots for users to chat with and ask various questions.
But the use of chatbots is not an exclusive trend in the tech world. As reported by Bloomberg, the CIA itself (yes, the U.S. intelligence agency) is also said to be developing an AI-powered chatbot similar to ChatGPT. The artificial intelligence will be trained on publicly available data, and alongside the responses, sources will be provided so that agents can verify their validity.
“We’ve gone from newspapers and radio, to newspapers and television, to newspapers and cable television, to basic Internet, to big data, and so on,” Randy Nixon, director of Open Source Enterprise at the CIA, told Bloomberg. “We have to find the needles in the haystack.”
According to Nixon, this tool will enable CIA agents and other U.S. intelligence agencies to search for information, ask follow-up questions, and summarize large amounts of data. Afterward, agents could directly pose questions to the AI to obtain answers on specific matters, as well as the source of the information.
But of course, when we talk about the CIA, the claim that this AI will only be trained on public data is at the very least… suspicious.
Nixon has stated that the tool will respect U.S. privacy laws; however, he does not explain how the U.S. government would protect it from leaks to the internet (as happened with Meta‘s LLaMA AI) or from the use of information obtained incompletely but technically “publicly.”