Recognising the potential and pitfalls of AI at work

Artificial intelligence is changing the way we live, but we need to develop the skills to make the best use of it, writes Will Rollinson

Anyone who has a Smarthub in their home has probably heard a phrase similar to “I’m sorry – I can’t help with that”. All the more alarming when you haven’t asked for any help!

There’s no doubting that artificial intelligence is changing the way we live. But it is also changing the way we work. Whilst there are opportunities from deploying technology, employers should be aware of the challenges AI can pose so that they can avoid being exposed to the risk of litigation, particularly in the areas of recruitment, the gig economy and discrimination.

AI can be particularly useful in sifting through applications and CVS. Software can search for relevant key words in minimal time, saving on intensive man hours. This could help to weed out unconscious bias which may reduce the risk of discrimination during the recruitment process. However, if the software is not programmed correctly, it could have its own built-in unconscious bias. This could potentially see a rise in discrimination risk. There are also data protection considerations as the General Data Protection Regulation has provisions which regulate automated decision making.

The much-talked-about gig economy has seen a litany of cases about employment status. It seems possible that technology could be very well placed to assess the true nature of someone’s employment status. This could help to reduce the risks in employment and tax litigation.

As the world of work becomes increasingly flexible, technology could be used to offer ad hoc projects to individuals, reducing the need for formal engagement. Technology could be deployed to identify workforce gaps and trigger recruitment processes to fill them.

There will inevitably be a lag between the way in which people are engaged and the way the law deals with them (either from a tax or legal rights perspective). This could increase risk if organisations do not continue to monitor legal developments.

Employment contracts and other documentation (including confidentiality, non-disclosure agreements, and restrictive covenants) will need to be looked at carefully to make sure an organisation’s interests are adequately protected when engaging staff in new and different ways.

Technology could help with the duty to make reasonable adjustments. Proof reading/editing software can already be used to help employees with dyslexia and staff with visibility problems can benefit from programmes which read text. As technology develops there is huge potential to overcome obstacles for disabled people in the workplace. This could massively improve equality and diversity and reduce the number of discrimination claims brought by employees.

Businesses must ensure the technology used is not itself discriminatory in any way. For example, any AI which requires employees to complete information to help assess their disability status and/or ways to help them remain in work/overcome any disability might not be well suited to certain individuals and may itself need to be adapted.

The technologies that exist today are scratching the surface of what is possible. In the world of employment there are many other areas that can and will benefit from AI – biometrics and remote working are areas where tech is already facilitating change.

And whilst reports of mass technological unemployment have been overplayed, perhaps the greatest challenge to employment is the potential for a skills gap to develop. It is therefore up to employers, employees and government to build a future workforce now that has the skills required to work in and adapt to AI in the workplace.

Will Rollinson is a senior solicitor in the employment team at Brodies