AI AND ALGORITHMS: THE RISKS OF OUR AUTOMATED WORLD
If the early part of this century was known for being the age of big data, then what comes next will be the age of the algorithms.
In the early 2000s, due to the rise of the internet and the considerable advances made in computing power, organisations were suddenly overwhelmed with huge volumes of data. In most industries, considerable investments have been made to harness the value in this data through advanced analytics, to inform decision making, enhance customer knowledge and personalise services.
However, arguably the most valuable characteristic of big data is its ability to power the development of artificial intelligence (AI) technologies. Machine learning (ML) algorithms are heavily reliant on big data. They are commonly trained through large data sets, which are used to represent the environment being modelled. This training data may be in the form of hundreds of thousands of images of people in the case of a facial recognition application, or similar numbers of sample voice recordings for a speech recognition system.
Such advancements are increasingly being utilised across industries to automate processes (fully or partially) and decision making. AI is used to power compliance tools in the financial services industry, assist with medical diagnoses in healthcare and undertake sophisticated targeted online advertising campaigns in consumer sectors.
Yet with new technologies come new risks. Algorithms can be inherently biased, resulting in discrimination against certain groups. The considerable reliance on data to power machine learning can pose privacy challenges, and autonomous machines may not perform as expected, resulting in damage to property or physical injury.
Apr-Jun 2021 Issue
Hogan Lovells International LLP