IntelCenter & Generative AI
Recent advances in Generative AI are changing how the world works and interacts with data. In some areas it is already bringing powerful new capabilities to bear and in others the day when it is ready to be deployed is rapidly approaching. Key to capitalizing on these opportunities is quality data and this is exactly where IntelCenter is uniquely positioned.
It is one thing to ask ChatGPT a national security question on a threat actor and get a reply based on the wild range of data found on the Internet and another to ask a Generative AI that same question on a high-quality curated set of structured data collected directly from those very threat actors along with supporting data created by a team of language-enabled analysts.
When you pair IntelCenter’s extensive targeted data holdings of more than 400 million data points stemming from more than 1.3 million records, an incredible array of opportunities begin to present themselves across analytical and operational areas. It goes beyond just quality data, however. In this rapidly evolving area you need true subject matter expertise to develop the tradecraft around what Generative AI can do today, what it cannot and where it is headed. It is critical to have humans in the loop with the proper expertise in that data and the tradecraft to understand responses and how to use them. IntelCenter, with more than 30 years of expertise in collecting, analyzing and exploiting data from the targets it works, is uniquely positioned to do exactly that.
IntelCenter has already deployed more than 40 different AIs in the IntelCenter Database (ICD) to do everything from transcription and translation to object detection and named entity recognition against threat actor generated content. This week IntelCenter will be rolling out additional AIs to perform topic extraction and summarization. In addition, IntelCenter will be also deploying its first AI to detect content generated by an AI with AI deepfake detection coming soon for both images and video. In all of these areas, IntelCenter leverages the best-in-class AI providers for each of its data types. It currently utilizes more than a half-dozen different providers to maximize what can be achieved today. In object detection alone, IntelCenter simultaneously uses five different AI providers and can rapidly onboard a new provider in just minutes.
Those existing AI services have already enabled IntelCenter users to interact with threat actor generated content like never before. A user can search across hundreds of thousands of images for a landmark or logo. Words on a road sign in Arabic can hit on an English keyword search, and numbers from a barcode on a weapon part can be found. Threat actor messaging searches can now identify something said in the middle of a video in another language instead of just the limited text appearing in a social media post.
Behind the scenes at IntelCenter, far more extensive work has been underway to create the foundation on which IntelCenter will begin to roll out its first Generative AI capabilities trained specifically on IntelCenter’s curated data to enable focused queries against specific threat actors and broader ones that look across multiple actors to find patterns and relationships that might otherwise not have been discoverable.
For IntelCenter, this incredibly exciting period is still anchored in a core principle that a human analyst, collector, targeter, etc. is absolutely essential to the process. When you combine that with the proper tradecraft and AI capabilities, great advances can be made in support of the mission. Even more significantly though, is that this is likely the only chance the national security community has at keeping up with an incredibly complex and rapidly evolving threat landscape as the opportunities AI presents are not exclusive to those working against the threats.
For further information on IntelCenter’s platforms and data holdings, please contact us by clicking here.