Information operations, or info-ops, are coordinated efforts to manipulate public opinion online. Think coordinated networks of accounts spreading misinformation or propaganda. These campaigns can be incredibly disruptive, undermining elections and eroding trust in institutions. But how can we uncover these hidden agendas? New research suggests that large language models (LLMs), like the tech behind ChatGPT, might be the key. Researchers used LLMs to analyze social media posts related to past elections and geopolitical events. They found that LLMs could identify the goals, tactics, and even the narrative frames used by these campaigns. For example, in the 2022 French election, LLMs identified coordinated accounts pushing pro-Ukraine messages and requesting assistance from France, potentially attempting to influence public opinion and government policy. The LLMs were also able to analyze posts related to the 2023 Balikatan U.S.-Philippines military exercise, identifying a coordinated campaign promoting a Bollywood movie. While seemingly unrelated, this highlights the diverse nature of coordinated campaigns and the potential for LLMs to identify spam or other forms of manipulation. The research also explored how these campaigns change their tactics over time, adapting to events and shifting narratives. By analyzing posts before, during, and after key events, the LLMs revealed how campaigns adjust their messaging to maximize impact. While this research shows promise, it also highlights the limitations of LLMs. They can sometimes misinterpret information or even "hallucinate" connections that don't exist. Human oversight is still crucial. However, LLMs offer a powerful new tool for understanding and exposing information operations, potentially helping us safeguard against online manipulation and protect the integrity of our information ecosystem.