Select a location

This selection will switch the site from presenting information primarily about Kenya to information primarily about . If you would like to switch back, you may use location selection options at the top of the page.


How employers can use Artificial Intelligence in data protection era

By John Ndirangu

Over the past few years, the world has witnessed increased use of artificial intelligence (AI) in almost every sphere of life. Some of the AI solutions such as ChatGPT by OpenAI have drawn immense public attention and are said to be “revolutionary”. Governments and businesses world-wide are exploring ways of integrating AI into the economy and businesses. The report of the Distributed Ledgers Technology and Artificial Intelligence Task Force which was formed by the then Cabinet Secretary, Ministry of Information, Communications & Technology, Mr. Joseph Mucheru in 2018 contains enriching recommendations on how the Government of Kenya can leverage on emerging technologies such as blockchain and AI to effectively deal with thorny issues such as corruption and unemployment and generally enhance public service delivery in areas such as healthcare, education, manufacturing housing and food security.

The current digital age dictates that everyone, including employers must be concerned about the opportunities and risks created by the emerging technologies. AI solutions are being used by employers in activities such as recruitment, performance management, data management and routine office functions. The technology is way ahead of legal developments in many countries but an employer would need to ensure that deployment of AI solutions in the workplace complies with the existing employment laws.

One of the key areas of compliance is the handling of personal data. Since AI thrives on the use of data, an employer using it must be conscious of the likely data protection issues. For instance, when in a recruitment process an employer uses an AI solution to record images and analyse the emotions and behaviour of candidates, the consent of each candidate would be required and the employer must ensure that there is no excessive or unnecessary infringement of the right to privacy. Employers would therefore need to formulate policies on who can use the AI tools and when, how and when to obtain consent from job applicants, how to inform them of their rights and that AI is being used, how to process and store the data received, what amounts to lawful processing of the data, when to conduct a data protection impact assessment, explain the significance and envisaged consequences of the processing of the personal data, among other legal obligations.

These restrictions on automated individual decision-making by a data controller or data processor are prescribed in the Data Protection Act, 2019 and its regulations which also require a data controller or data processor to ensure the prevention of errors, use appropriate mathematical or statistical procedures, put appropriate technical and organisational measures in place to correct inaccuracies and minimise the risk of errors, among others. An employer must therefore understand both the AI software and the surrounding legal environment to eliminate or at least mitigate potential legal risks.

Both the Employment Act, 2007 and the DPA require fairness in employment and labour practices. Some of the principles of fairness include granting the data subjects the highest degree of autonomy with respect to control over their personal data, elimination of any discrimination against a data subject and incorporating human intervention to minimize biases that automated decision-making processes may create. AI outputs may create bias if the data used is not objective or is based on certain biased assumptions.

Legal issues may also arise from the application of work injury compensation laws where injuries result from employees’ interaction with AI-powered machines and systems. The use of AI may also mean that an employer has to reduce the workforce through redundancy, which is a highly regulated and costly procedure under Kenyan law. There might also be internal re-alignment of changes to roles held by retained employees regarding their job description and remuneration. Trade unions may also oppose changes especially those which may result in job losses. These changes may affect the employees’ mental well-being, morale and performance. The legal risks which may arise are unlimited. One of the ways to prepare well for handling some of the issues is relooking at the existing human resource policies and refining them, if necessary. It is also necessary to train staff to on corporate policies to address any gaps in technical skills.

There are no universally accepted laws yet specifically regulating the use of AI but various countries world-wide are coming up with laws and policies to address the existing gaps. In April 2021, the European Commission published a proposal on AI regulation which, if enacted into law, will provide guidance to EU members on the kind of regulation to adopt. The Proposal contains proposed rules on the design and use of some high-risk AI systems in employment management, algorithmic discrimination and prohibited AI practices. The passing of new laws shall require entities using AI solutions to conduct assessments of their internal systems and procedures to ensure compliance.

Like any other regulated area, the use of AI comes with legal consequences for non-compliance with the law including fines. In August 2021, the Italian Data Protection Authority fined a food delivery company EUR 2.5 million for the unlawful use of algorithms for the management of its riders and for not being transparent about the use of the algorithms. In July 2021, another company was fined EUR 2.6 million for almost similar violations. Sanctions by regulators may also have an adverse effect on the reputation of a business. Various policymakers recommend the use of AI systems that meet key requirements such as privacy and data governance, transparency, accuracy, non-discrimination, fairness and diversity.

Time is ripe for Kenyan businesses to think through how AI impacts or will impact them in future and take the necessary corrective or preparatory actions. For employers, committing resources to the development of the workforce in order to gain the right skills and knowledge is not an option, if AI is to be embraced successfully in the workplace. Staying abreast with legal developments in this space is also advised. Employees, on their part, must be ready to learn new skills and adapt to change. Everyone must really be willing to tap into the full potential of this exciting and evolving area.

The article was published in the Business Daily on 26 May 2023 and can be accessed here.