Consultation is key where employee data is concerned
By Naemi Groh
People analytic tools use the power of data to help companies improve how they identify, attract, develop and retain talent. But while the benefits are enticing, such systems also raise ethical concerns, writes Naemi Groh
Algorithms are deployed in a wide array of technologies that are becoming more important in our daily lives, from robotics to self-driving cars and medical diagnostics. The world of work is also increasingly data-driven, with HR departments no exception. Advances in technology and AI mean that increasingly complex algorithms are being used to solve classic HR issues, with ‘people analytics’ – the application of digital tools and algorithms to people-related data – one of the hottest buzzwords in the industry. The aim of people analytics is to extract ‘actionable insights’ that improve people management processes and enhance operational capabilities (for example through increased efficiencies, cost reductions and risk mitigation).
As interest in people analytics has grown, so the number of tools to filter, connect and analyse data has risen proportionally. Systems have been developed which – with the help of AI – can analyse patterns in spoken and written language in a way that can assess an individual’s ability to communicate as well as other competencies and signs about their character. Speech analysis tools are predominately used in recruiting, where applicants may be asked to take part in an automated telephone interview with 15 minutes of random questions asked by a computer. The interviewees’ answers are then analysed against around 5,000 voice samples to prepare a personality profile.
Another tool analyses communications data in real-time to understand how people work and to uncover patterns that can help businesses make better decisions around organisational health, workplace strategy and process optimisation. Employees carry a ‘sociometric badge’ which measures the frequency and duration of face-to-face interactions and tracks their movements. Combining this with other e-mail and chat communications data, as well as information on, for instance, lightbulb usage, companies can optimise communications between different divisions and better manage their space requirements.
However despite these tempting possibilities, organisations should be aware not only of the legal risks associated with the introduction and use of people analytics tools (for more information see our recent briefing), but also of the ethical dimension of people analytics. Many people find the idea of being analysed or judged by computers unnerving, and there has been media coverage of employers using big data collected by outside firms to track pregnancies or Amazon developing wireless wristbands to track warehouse workers.
A business looking to introduce people analytics solutions is likely to find implementation easier if it holds a proper, transparent dialogue with its workforce to allay any fears and to establish trust in their use. The ethical dimension of AI and how it is influencing the world of work is also likely to be high on the agenda of unions. UNI Global Union, the global union federation for skills and services, has issued a set of principles (‘Top 10 principles for ethical artificial intelligence’) which it urges other representative bodies, shop stewards and global alliances to implement in their collective agreements, global framework agreements and multinational alliances. These principles state that workers should be consulted on the implementation, development and deployment of AI systems. They also provide that workers have the right to understand exactly how and why an AI system has made a decision, and the right to appeal it. Other recommendations discussed by people analytics experts include the establishment of a governance council or the publication of a code of practice which sets out how a company will handle employee data, which might even be co-created together with employees.
When navigating the fast-changing world of advanced analytics, companies should always consider the wider implications or risk serious resistance from employees or works councils, as well as reputational damage. Essentially, companies have to find the right balance between what they could do with people analytics tools and what they should do.