The Evolutionary AI Race is On
Autonomous vehicles, competing in strategic games, image recognition, online assistants, predictive maintenance, object detection and classification are just a few of the incredible, positive ways in which AI is shaping our future.
AI simulates human-like abilities such as hearing, seeing, reasoning and learning and has applications across almost every industry. AI includes multiple technologies, such as deep learning, computer vision, natural language processing and machine reasoning. Any process that perceives its environment and takes actions to maximise its chances of successfully achieving its goals is classified as AI.
For most, these goals would be things like improving efficiencies through predicting manufacturing faults, enhancing customer experience by employing behavioural adaptation to improve the emotional intelligence of customer support personnel, providing highly accurate film and music recommendations and helping to solve society’s toughest challenges. But for criminals, the goals are obviously more nefarious.
Let’s look at some likely use cases of cybercrime underpinned by AI:
Hacking personal accounts - through mining huge amounts of personal data that lie in the public domain such as our social network data, location, birth, gender, telephone numbers and email addresses
Personalised phishing emails - monitoring emails and text messages to generate highly personalised, and increasingly deceptive, emails
Impersonation through voice synthesis - gathering social network voice data and impersonating people to commit fraud
Automated hacking and cyberattacks - the use of AI to automate tasks involved in hacking or executing cyberattacks could mean attacks on a much larger scale than previously seen
Physical Interception - using AI to automate tasks involved in carrying out attacks and subverting cyber-physical systems like drones to cause crashes or alter targets
Manipulating public opinion and news through the use of highly believable fake videos and bots
There’s no doubt that AI will be used by those with malicious intent. As AI matures and becomes more of a commodity, criminals will become increasingly capable of targeting vulnerable users and systems. Security firms and agencies will likewise lean on AI in a continuous effort to keep up. Some feel that AI technology should be more regulated. Considering the broad range of technology that falls under the umbrella of AI, this could be tough. There’s also the risk that restricted access to AI technology would hinder the positive impacts and potential of AI to transform our lives.
A report titled The Malicious Use of Artificial Intelligence has just been published by 26 researchers from universities and organisations in the field of AI. Read it. Accept that AI “is a dual-use technology”. And seriously consider how we can put best practices in place to mitigate harmful applications of AI.