Investigating AI as a Consumer Platform: Part of the OSINT Toolkit
AI deserves a place in the OSINT toolkit, not because it's some magical new technology, but because AI platforms are increasingly becoming online environments rather than simple tools.
As investigators, we're already comfortable examining social media accounts, GitHub repositories, Discord servers, forums, and countless other online spaces because people leave traces of their interests, relationships, activities, and identities there. AI platforms are beginning to follow the same pattern. Users create profiles, publish custom assistants, share prompts, build research collections, and participate in communities that are forming around these platforms.
The fact that AI platforms haven't yet generated a long list of landmark investigative cases isn't particularly important. We've seen this cycle before. New technologies often start as specialized utilities and eventually evolve into ecosystems where people create content, build reputations, establish connections, and leave behind artifacts. Researchers examining public AI ecosystems such as OpenAI's GPT Store have already shown that thousands of user-created assistants can be analyzed at scale, revealing creator information, websites, areas of expertise, and other publicly visible signals. Whether these platforms ultimately become major OSINT sources remains to be seen, but they are clearly becoming environments that investigators should understand.
We don't need to speculate about whether AI-related artifacts might someday become evidence. They're already showing up in courtrooms.
In one recent Florida homicide case, prosecutors introduced ChatGPT conversations as evidence related to planning and intent. In separate civil litigation, federal courts ordered the preservation and production of millions of ChatGPT conversations during discovery. What is significant about these cases is not the artificial intelligence aspect. It is the digital footprint aspect. Every time people adopt a new platform, they leave traces behind. AI platforms are increasingly generating the same kinds of artifacts investigators have examined for years on social media sites, forums, messaging applications, and search engines.
Sources:
https://theportlandmedium.com/news/national-news/ai-conversations-courtroom-implications/
One of the most interesting investigative applications may be attribution. Determining who created content, controlled an account, or carried out a particular activity is often more challenging than finding the information itself. AI-related artifacts can provide additional attribution signals that help fill in those gaps.
Public profiles, usernames, custom AI assistants, image-generation galleries, prompt libraries, research collections, and even behavioral patterns can all contribute to a broader attribution assessment. None of these indicators should be treated as definitive on their own, but when combined with other sources, they can help strengthen confidence in an assessment.
AI artifacts can also serve as valuable pivot points. A username discovered on a public AI platform may appear on GitHub, Reddit, Discord, LinkedIn, or X. A custom AI assistant might link back to a business website. A shared research project could reveal professional interests, technical expertise, or organizational affiliations. As with many forms of OSINT, the real value often isn't the initial discovery. It is the connections that emerge when information is correlated across multiple sources.
At its core, this isn't really a story about artificial intelligence. It's a story we've seen repeatedly throughout the history of the internet. People adopt new platforms. They create content. They establish identities. They leave artifacts behind. The platforms may change, but the investigative principles remain largely the same.
For investigative capabilities, see our draft AI investigation guide in our Resources section.
