How AI Is Paving the Way for Improved Surveillance and Cybersecurity

By on
Read more about author Martin Ostrovsky.

Cybersecurity is increasingly leaning towards artificial intelligence (AI) to help mitigate threats because of the innate ability AI has to turn big data into actionable insights. Rightly so, because the threat to data security is real, and across all industries. For instance, while there were fewer than 50 million unique malware cases in 2010, the number had risen to more than 900 million malicious executables in 2019, per the statistics of the AV-TEST Institute. Another report states that malware is the most concerning cyberthreat targeting organizations, with phishing and ransomware jointly ranked second.

The complexity of cyberthreats has increased as well, even as cybersecurity companies reimagine their product portfolios. With the explosion of the Internet of Things (IoT), this challenge is only going to get bigger. Modern-day solutions that include not just our smartphones and the mobile apps that we use, but also smart devices installed in our homes or vehicles, are perfect examples of how breaches can happen via malware and ransomware. Our GPS, Google photos, smart thermostats, and smart light switches all store information that gives away indicators of our daily routines. This data is a goldmine for cybercriminals. All it takes is one device and the malware obtains access to the whole network system, which in turn means that protecting a single-entry point is not enough anymore. 

The good news is that cyber surveillance organizations and government agencies are cognizant of this phenomenon. There has also been a significant increase in the demand for cloud-based security solutions among small and medium-sized businesses (SMBs). All this has resulted in driving the growth of artificial intelligence in cybersecurity. According to one report, AI in cybersecurity is anticipated to reach $46.3 billion by 2027. This number is expected to grow even more given the sudden shift of countless businesses towards digitalization due to market disruption caused by the COVID-19 pandemic.

AI, Machine Learning, and Data Security

AI capabilities are rooted in machine learning (ML) tasks such as natural language processing (NLP) as well as applications like graphical processing units (GPUs) for 3D data or Google’s own application-specific integrated circuits tensor processing units (TPUs) to accelerate machine learning workloads. These powerful tools help train complex models of neural networks as they discover trends and patterns and trigger actions in text and video data to detect security risks.

Text and Video Analytics

AI software collects a large amount of security event data from different sources and analyzes it using text analytics and background modeling for videos. It accesses sources like social media comments, user-generated videos on accounts like TikTok, Facebook, tweets, phone messages, etc. and identify anything that is an anomaly. It also compiles incident reports. This facet is used increasingly by law and order agencies to mitigate national security threats, child endangerment, help in suicide prevention, and other critical areas.


Machine learning algorithms also enable a secure network to users along with providing crucial alerts of data breaches. AI-based blockchain technology is an example. Blockchain allows secured visibility combined with complete transparency, wherein AI works with huge volumes of secure data and gives useful insights through machine learning techniques. As a result, several people have secured access to a chain of data tracks, thus allowing for a trusted platform to store critical information. This is why the technology has become the bedrock in banking and supply chain management systems.

AI in Surveillance and Cybersecurity

AI is being implemented in surveillance and cybersecurity in mainly two ways:

1. Text analysis and incident reporting

AI algorithms gather data across numerous sources including social media, chat forums, and cell phone and app messages to detect cyber threats or vulnerabilities. Natural language processing tasks further identify specific keywords, extract them from whichever source they occur in, compile, and summarize it. These algorithms can also gather information on the origin of the text, the latitude, and longitude, as well as the IP address of the user. 

Intelligence Reports

NLP also enables AI programs to generate automated cyber threat intelligence reports (CTI) that can give early indicators and warning signs of unusual activities on a given network. Reports like these have helped financial institutions tremendously in mitigating fraud and thefts – so also industries like hospitality and healthcare.

Incident Diagnosis Reports

AI examines past data to see patterns and anomaly indicators in the network activities that can tell data scientists the root cause of an incident. Once the cause is identified, predictive and prescriptive analysis can be used to contain the issues that lead to the root cause and help parties take corrective actions. This can include improving functions like bettering the quality of cyberthreat intelligence data by adopting new data sources, improved diagnostic processes, and recalibrating reporting.

2. Video content analysis

Powerful ML algorithms can analyze videos for their content by converting audio to text and extracting any topics or words that have been deemed dangerous to the public. Video content analysis, importantly, also includes identification and extraction of background imagery, logos, objects, and any other key features that can point to anything that is a threat to the public. 

Video analysis is used to detect not just threats to security promulgated by terrorist organizations but also those that are spread by way of misinformation that can cause great damage to society or create chaos in governance. The recent example of conspiracy theorists taking to social media and spreading misinformation on COVID-19 lockdowns and targeting governments, as well as the anti-vax movements, are lucid examples of how cyberspace can be used by anyone for vicious activities.

With the help of ML and NLP, AI platforms can read and analyze multiple languages without the need for translations. This leads to a higher degree of accuracy in detecting, extracting, and analyzing phrases, words, and topics based on cultural and language semantics, leading to lesser false alarms. Similar to text analytics, these intelligent AI algorithms continually learn from video content and in the process get smarter and more accurate over time.

Apart from this, AI engines also detect anomalies in movements, identify people through facial recognition, and know the difference between objects and humans. This is key in identifying threats either by persons or unidentified objects left in public places. 

  • Facial recognition: Facial recognition is crucial in crowd control and hazard safety, as well as for law enforcement agencies to identify potential threats. In recent times, facial recognition has been used to monitor the use of face masks for COVID-related health and safety measures.
  • Motion detection: Through AI, deep learning algorithms analyze terabytes of video data to track moving objects in 3D. ML-based surveillance software can identify a human from a moving object in the background, and differentiate between objects in proximity. AI-driven video surveillance technology can also connect to existing and outdated CCTV infrastructure to enable motion detection and analysis. 


We have to adapt to a new reality where data security, and cyberspace itself, are becoming more and more unpredictable and complex. We cannot wish away these challenges. That’s why we need to have continual research and considerable investment in cybersecurity Data Science in order to develop even smarter, faster, and sharper solutions for public and data safety.

Cybersecurity companies are already considering this crisis as a new opportunity to think out of the box. Added to this, practices like increased cross-organizational data security trainings, stronger implications for companies who don’t follow legal data privacy obligations, and tax incentives for companies who adopt advanced AI-based solutions can help in cybersecurity and surveillance.

Leave a Reply

We use technologies such as cookies to understand how you use our site and to provide a better user experience. This includes personalizing content, using analytics and improving site operations. We may share your information about your use of our site with third parties in accordance with our Privacy Policy. You can change your cookie settings as described here at any time, but parts of our site may not function correctly without them. By continuing to use our site, you agree that we can save cookies on your device, unless you have disabled cookies.
I Accept