Our primary objective with the protective intelligence blog is to share ideas and create a discourse that will help like-minded security professionals protect assets and ultimately save lives. As talented as our team of contributors is, we see great value in highlighting important works created by our colleagues in the industry.

There are countless examples of artificial intelligence, machine learning, and automation slowly becoming a part of our day to day activities – the investigative process is no exception. For that reason, we are thankful for the opportunity to share an article from one of our experienced industry colleagues. In this article, Julie Clegg (Founder & CEO at Human-i Intelligence Services Inc.) gives us her perspective on how artificial intelligence and machine learning are empowering investigators by streamlining mundane tasks, and freeing analysts to focus their energy on creative solutions to investigative challenges. We hope this forward thinking article provides you valuable insight about how our investigative work will be influenced in the future.

(Originally published on July 6, 2019)

CREDIT: “The Impact of AI & Machine Learning on Investigations” was originally produced by Julie Clegg via World-Class Investigator™.

From computer-assisted medical diagnoses, to financial trading, to market personalization; artificial intelligence (AI) is becoming increasingly relevant, and prevalent, across most major industries. As the technology develops and is refined, often via sophisticated machine learning algorithms, more innovative applications are discovered to increase efficiency, reduce costs and supplement or replace human capability. Not only is AI used to engage with and respond to customers in millions of online and voice interactions each day, these computers can learn to anticipate and solve problems, direct enquires, and respond to complex instructions and combinations of scenarios as they arise.

Anyone using Alexa, Siri or Google Home is interacting with AI every day in the form of “Internet of Things” connected appliances, such as gaming consoles and myriad available smart devices that have been programmed or “trained” to respond to verbal or physical triggers or commands with pre-programmed actions or information.

In investigations, internet-connected tools that utilize artificial intelligence make information gathering and analysis less expensive and time consuming while proving to be more accurate and often, more covert, reducing risk and increasing solve-rates.

As AI and machine learning continue to accelerate the processing of data at rates of speed and accuracy that are significantly superior to humans, and data retrieval, analysis, and dissemination become increasingly automated, investigators are voicing concerns at the possibility that we are approaching an era whereby human investigators are no longer needed at all.

How AI and Machine Learning Are Used in Investigations

Timely, accurate and verifiable evidence is critical to the success of any investigation and there can be little argument about the practical application of AI technology in most cases with any digital or technical elements.

AI tools allow investigators to spend their time collecting and evaluating evidence that drives the investigation forward, rather than sifting through data. The laborious task of manually extracting data, if it even exists in a format that allows it to be examined or extracted, now takes a fraction of the time, and reduces the risk of critical errors and omissions.

AI with machine-learning capability can be programmed to not only extract data but to identify, evaluate and classify it while excluding or flagging that which is irrelevant or not based in fact. Deep learning and neural networks represent powerful machine learning-based techniques; in forensics, text mining tools can conduct in-depth linguistic analysis of digital documents such as emails, surveys, blogs, and reports.

The Future of AI in Intelligence Gathering

As recently as a decade ago, voice printing and facial recognition was the stuff of science-fiction and spy films. Today, AI-assisted surveillance devices regularly scan, record, and analyze live digital footage to identify people in crowds, at borders and in the workplace.

Surveillance cameras supercharged with AI not only capture incriminating footage with their digital eyes but are smart enough to analyze it and “decide” to act – from activating a warning system to notifying the authorities. AI is also routinely used to speed up the evidence review process by scanning, tagging, and highlighting still frames in videos at speeds far superior to a human being.

From the FBI’s facial recognition database of hundreds of millions of faces to the CIA’s plan to replace spies with AI, it’s clear that machine learning and AI will play an increasingly significant role in both physical and virtual security and investigations in the coming years. In 2016, Barack Obama’s White House released Preparing for the Future of Artificial Intelligence, a white paper discussing the government’s recommendations for increasing the use of artificial intelligence, and in 2017, Robert Cardillo, the director of the National Geospatial-Intelligence Agency (NGA) announced his push toward “automation” in which he envisioned a future in which robots would perform 75% of tasks currently undertaken by human intelligence analysts.

As humans increasingly interact with technology both passively via surveillance and routine data collection, and actively via social media and smart devices, exponential quantities of unsecured data is uploaded, shared and analyzed every day.

Social media generates billions of digital conversations every minute, giving businesses and governments insight into consumer behavior and preferences, as well as national and international risks and threats, knowledge gaps and technical vulnerabilities.

From social listening in marketing, competitive intelligence in business, behavioral and social profiling, threat assessment and early warning systems in the security industry, augmented intelligence is a gateway to an unprecedented level of power and control over our own data, and that of others. Investigators have never had so much verifiable intelligence at our fingertips, so cost effectively and at such speed, and yet, it is easy to become confused and overwhelmed by the technology, it’s effectiveness and it’s trustworthiness.

As exponentially as the available data increases, so too do the tools for processing said data, and often, the complexity of their implementation. For investigators unfamiliar with technology, or who are not routinely engaged in digital cases, the prospect of keeping up with AI-enabled tools and machine learning can be daunting and is often avoided entirely, resulting in mis-aligned expectations and lost opportunities for locating accessible information.

While the future is, without a doubt, technically driven, traditional investigation skills and techniques are more important than ever to ensure that the data is contextualized, and vital, non-digital information is not overlooked or omitted. A successful investigation in the 21st century must include all-source data; it is no longer possible for investigators to be skilled in every investigative element. Collaboration is more important then ever, as specialisms narrow and the scope and scale of data expands. Traditional investigation skills may become rare and thus, more valuable, as digital natives focus on the technical aspects while the more “human” skills that determine motive and motivation lose focus.

Conclusion

From eliminating time-consuming tasks to sorting through massive amounts of raw data, the most significant benefit of AI in investigations is its ability to speed up the processing of data, while improving accuracy and reducing costs. But while AI tools are vital in today’s digital environment, they are not likely to replace human investigators now or in the future. AI may be able to use its automated algorithms to calculate and predict patterns in data, and even human behavior; however, only another human can truly comprehend and relate to human nature with the kind of nuanced discernment and context required in complex investigations. Crime is ultimately initiated by humans, as a result of unique motives and motivations, and even the most advanced AI algorithm cannot entirely process random, unpredictable, irrational or emotional events.

Cardillo, the NGA director, has stressed that AI isn’t meant to replace analysts but to “elevate” them and help them achieve more. In investigative work, the rise of AI tools doesn’t mean a robot takeover that renders humans obsolete; it means investigators will have more time to focus on the meaningful work that ultimately closes cases, supported and improved by artificial intelligence. The future of intelligence in investigations is augmented, not artificial.


About the Author

Julie Clegg, Founder & CEO at Human-i Intelligence Services Inc.

From the computer screen to the television screen, Julie has crafted her life’s work around investigation, profiling and technology, combining her knowledge of cyber-psychology with two decades of online investigations experience to bring to light those that hide in the digital shadows and cause disruption, distress or terror.

A former police detective and undercover operator, criminal and geographic profiler, an active search and rescue volunteer, a passionate human rights advocate, a global educator and an avid world traveler, Julie approaches her life and work with determination and a relentless sense of adventure. She is driven by a desire to help every person live a life free of victimization, oppression, and fear.

Along with her trusted team of leading global experts in digital and mobile forensics, data analytics, forensic psychology, and organized crime, Julie uses all-source intelligence with a focus on online OSINT, social media intelligence, and ethical and proportionate practices for covert and undercover investigations to solve complex problems for corporations, governments, law enforcement and non-profits worldwide.

In a world of increasingly artificial intelligence, Julie brings human solutions to digital investigations.

About The Protective Intelligence Blog

By every metric, the role of protective intelligence is growing increasingly important for your security program, as it operates domestically and (especially) internationally.  Protective Intelligence is our medium for understanding not only threat matrix and risk level, but also trends, problems, solutions, as well as ideas to support the mission of protective security professionals.  The speed by which we can send and receive information, and the amount of information we need to evaluate, has eliminated problems in some areas and exponentially compounded problems in others.  Our team seeks to address issues stemming from these problem areas, offering next-level analyses and proven solutions with an eye toward the future.

Our content contributors come from organizations involved in protective intelligence research, corporate executive protection, threat assessment investigations, and related security intelligence fields.

Subscribe to Protective Intelligence Thought Leadership