As cybercriminals become increasingly sophisticated with their scams, using artificial intelligence (AI) to scale up their operations, it can feel like fighting a losing battle. However, AI and machine learning (ML) can also be used against them. Here’s how:

  • Detecting malware. Cybercriminals often try to hide malicious code by burying it in innocuous-looking code. AI is able to identify this by quickly cross-referencing against huge volumes of past code to find the slightest difference.
  • Preventing complex attacks.  ML algorithms, over time, can learn to recognise potentially complex assaults on systems. They could even prevent future attacks by identifying the cybercriminal and taking action against them
  • Detecting malicious bots. Bots make up a high percentage of internet traffic, and some can be dangerous. AI can quickly detect these malicious bots and take action to prevent them from doing harm to the network.
  • Tackling phishing scams. As phishing emails become more sophisticated and harder to identify, AI can be used to cross-reference against historical data of phishing and non-phishing emails. It will determine whether an email is legitimate or not.
  • Monitoring network traffic. AI can also be used to monitor network traffic patterns and recommend security rules – something that would normally be done manually. This frees up security professionals to work on other tasks.

Scamming the scammer

Some people use AI to go further than simply detecting and preventing cyber attacks. A New Zealand cybersecurity company uses AI to detect phishing emails, then creates a proxy email to reply to the sender referencing the scam attempt. This wastes the scammer’s time until they realise they’re emailing a bot. 

This can also be done with scam phone calls. Speech recognition AI can determine what a scammer is saying and respond with relevant stock answers mimicking a real person. Again, the aim is to waste scammers’ time. Eventually, ML could learn to respond more naturally, thus making it harder for the scammer to detect they’re talking to a bot.  

As the scammers become more sophisticated, it requires a sophisticated response. AI could be the solution.