Artificial Intelligence (AI) is leaping out of the pages of science fiction and into our daily lives at a rapid rate. Technology is transforming the way businesses operate, and many are looking to AI to support their change and growth plans.
AI essentially mimics human learning abilities and can be used to replace previously human processes. An example of this being put into practice are Chatbots that are now being used by many businesses to interact with the public and even recruit new employees. But if AI is going to be used in businesses decision making processes, how could it affect the decisions made?
Recent articles in the media have reported that some AI algorithms can resemble or adopt human bias so it’s vital for businesses using AI for recruitment purposes to understand how their program operates, some things to look at include:
- Diversity: AI will be built with the ethics, morals, biases and access of the people who create it, but who is creating it? This will have a significant effect on how AI makes decisions. If you are going to use AI this is something to explore further.
- Interaction: As AI learns through interaction it can be taught the biases of those around it, known as emergent bias. When machines are taught to learn from those around them, they don’t decide which things to filter but accept it all and adapt accordingly.
- Data-driven bias: Machines do not question the data they are given — they look for patterns within it. As such when the data that is fed to AI has already been skewed by human bias, the AI results will in turn be skewed, as well.
- Similarity bias: This emerges when algorithms distort what a machine will show people when they search for information online. As opposed to showing them all the options, it shows them the options it believes they are most likely to want or agree with.
According to McKinsey, tech companies spent somewhere between $20 billion and $30 billion on AI, mostly in research and development (R&D), and with over half of multinational businesses prioritising R&D and tech investment in 2018 (according to our Autumn 2017 Risk and Confidence Survey), this spend is likely to increase.
It’s no surprise then, that the inner workings of their programmes are confidential and commercially sensitive; however it’s important for businesses using the AI technology to understand potential issues surrounding bias. If implemented correctly, this AI technology could minimise the potential for human bias, but if organisations aren’t careful, it could exacerbate the issue or bring new and unknown challenges.
Matt Sumpter, UK Underwriting Director - Technology & Cyber Risks
Download our Risk and Confidence Survey at http://www.cnahardy.com/pulse