AI is disrupting agencies without the right infrastructure
It’s important that we make sure we don’t let our imaginations carry us away. AI will not give us abilities that we did not have before. This technology will allow us to do things we already know more efficiently. As with any emerging technology, if an organization’s foundations are not fundamentally sound, AI will disrupt operationsnot improve them. So, AI provides us with a great opportunity to ensure our foundations are strong and verify that we have done the careful work required to realize our goals and aspirations.
To use AI correctly, we need to understand what we want from it. generative AI model. We feed the model with data, and each human working with the model will add to that data. If you have 10 or 20 people working with an AI model, you have an incredible human engineering problem. We need to understand what we want the model to do, but also understand how we influence human behavior when we interact with the model. We need to ensure that we have the appropriate controls in place to allow the model to access the right data to perform its functions.
State Department says AI holds great promise for trend analysis threat intelligence. For example, we have seen white papers written by international students on how to crack encryption or how to exploit certain elements of operating systems for nefarious purposes. A threat of this nature will not generally appear in an intelligence report. But we want to know if academia is developing such malicious activities. We want to watch for any behavior like this that we haven’t seen before and develop defenses against it.
DISCOVER: There are many ways to ensure interoperability between Zero Trust tools.
Reducing technical barriers to entry for cyber analysts
We also want to lower the technical barrier to entry for cyber practitioners. In our security operations centers, we have a lot of talented people who may not have a high level of technical expertise. Yet they are valuable minds when it comes to developing cyber defenses. They may be able to use AI models to explore avenues of defense against specific attacks.
Our office witnessed an engagement with an advanced adversary, and an analyst realized that specific behavior was occurring at the same time each week. This information was vital for analyzing trends in this threat situation. We could use this information in an AI trend analysis to quickly identify other instances that fit the pattern.
As Foreign Service Officers, we change jobs every two or three years. We will assign a political or economic leader to a country and they will serve there for two years. The agent will make many useful contacts and collect good information. Then, after two years, the officer leaves. It may take a gap of two months before another agent arrives to replace him. Our Analytics Center is exploring ways to use machine learning and AI to enable new officers to quickly bring all of their predecessors’ work into context so they can quickly stay up to date on the situation of United States in this bilateral relationship. This is a simple but great use case where we can do something faster and more efficient.
RELATED: ModelOps helps agencies innovate in artificial intelligence.
Bring everyone to the table
We must hold AI to the same rigorous standards as all applications. We need to monitor what data is accessed by an AI model, and we need to monitor which people are accessing which AI model has access to that data. Administrators must map an ecosystem to understand what normal behavior looks like so they can understand abnormal behavior.
We also need to be consistent. Large language models act as large probability engines. They are trying to determine what is the most likely thing that will happen next. This leads to inconsistencies, so there must always be a human to verify the results. You can ask a large language model the same question and get a slightly different answer each time. Thus, humans must ensure consistency in how we view a problem and how thinking guides our responses. For the model to be consistent, we need to build processes in which humans continue to make decisions.
The advancements and applications of AI present a great opportunity to bring together more people who may not have been there before and allow them to share their ideas. AI, like any emerging technology, allows us to get back to basics and ensure our fundamentals and foundations are strong. AI has captured everyone’s imagination. This gave us a great opportunity to engage with segments of our own organization that we may not have collaborated with before and hear new ideas. While this may not always result in a technical solution, it can result in a stronger foundation within our organization.
FOLLOWING : Hyperconverged infrastructure supports AI for faster data analysis.
