Advanced technologies have always been vital for maintaining public safety and giving agencies a significant edge. But the last couple of years have seen an explosion of artificial intelligence capabilities, fundamentally changing the game for every area of security – from counterterrorism and intelligence to law enforcement and border control.
At the heart of this revolution are two developments. The first is the exponential growth of publicly available unstructured data, such as text, images, video and audio. Unstructured data is today the world’s largest data repository, growing in volume and significance with every passing day.
The second development has to do with groundbreaking advancements in AI capabilities, especially in the areas of Artificial Neural Networks and Deep Learning. These subsets of Machine Learning have been consistently delivering results far better than expected even by their own developers, pushing forward computational abilities such as pattern recognition, image recognition, natural language processing and machine translation.
Yielding these AI abilities to the intricate mesh of the unstructured data which makes up cyberspace reveals a hidden layer of reality. Hidden relationships surface, clues are found in immense image repositories, and intent is exposed. AI platforms are empowering the public sector to accelerate investigations, get conclusive answers to the most pressing questions, and uncover deep, actionable insights from the vast and growing ocean of unstructured data. Here are some examples.
Case 1: Unveiling hidden relationships
In a recent high-profile investigation, a murder suspect was known to the authorities but had an ironclad alibi. When the investigation reached a dead end, investigators turned to visual data from numerous sources to uncover connections between the suspect and other individuals who could shed light on the case.
By applying visual analysis to voluminous datasets, AI demonstrated how the suspect with the alibi was connected to the others through deep analysis of visual cues. Exposing the relationships between the original suspect, his connections, and the victim led to the eventual arrest of a new suspect.
The unearthing of hidden relationships between criminal group members spread around the globe has many potential applications relevant to those in the homeland security field, including in combating terrorism; human, arms and drug trafficking; and more.
Case 2: Securing a public figure
Consider the following scenario: In a major European city, public safety agencies cooperate to protect a visiting public figure. Amidst escalating threats against the dignitary and his entourage, concerns are raised regarding his security in a volatile, challenging urban landscape.
AI, and specifically cognitive computer vision and pattern recognition capabilities, help tackle this challenge by interpreting the behavior and relationships of and between suspect individuals captured in video and security camera footage. By applying cognitive analytics to voluminous visual data, AI creates an additional level of automated insights, exposing adverse behavior patterns and connections in real time. This augments the ability of agents on the ground to instantly pinpoint anomalous activities and persons demonstrating suspicious behavior and affinity within the public figure’s vicinity.
Case 3: Identifying connections to extremist organizations
At government facilities in the United States and abroad, situational awareness is critical for securing military posts, managing interactions with local workers who provide on-site services, and preventing hostile infiltration.
By analyzing and understanding extremist content, AI can turn up narratives built around themes of interest. This flags key figures and influencers – including non-obvious ones – who may be connected to extremist organizations or displaying extremist tendencies. Unveiling this information to those charged with securing government facilities enhances the intelligence assessments needed to ensure security.
Case 4: Securing airports
With tens of millions of international arrivals and departures yearly, the potential for security breaches at airports cannot be ignored. Just one person with malicious intent acting on their extremist agenda can have an outsized impact on public safety and passengers’ sense of personal security.
AI can be of tremendous value here, helping airport security personnel evaluate passenger lists, perform quick behavior/affinity analyses, and flag risks before they are realized. This ability to pinpoint risk can help ensure a smoother passenger travel experience – and might just reduce wait times in security lines along the way.
Case 5: Uncovering immediate insights in voluminous documents
The prototypical example of an unstructured data type is a large text document. Analysts are often tasked with absorbing these types of documents in full. But if the key information known or suspected to be contained in a document cannot be located, the document has little value. Using AI to parse this information can help analysts save the days or even weeks traditionally spent poring through thousands of pages of unstructured data in search of relevant information by making the data easily and intuitively searchable and filterable.
How AI will change what we do
AI is the facilitator that can help us make sense of a world of data that is growing exponentially all around us. We’re on the cusp of a world where evildoers are flagged earlier and criminals are caught quicker – or can even be deterred from committing crimes by the knowledge that they will be caught. Down the road, we may see new jobs created, implementing missions whose essence will be determined by new information that AI brings to light. This technology can be a powerful ally to those entrusted with national security and public safety, and we’re optimistic that the partnership between human and machine will yield results that outpace what we can imagine.
A version of this article has been previously published in Homeland Security Today
