AI is about to change the world of work. As an expert in Performance Management and Change Management, I’m fascinated by the possibilities of AI:
– opportunities arising from incorporating AI into performance management
– the risks and how we can mitigate them
𝐎𝐩𝐩𝐨𝐫𝐭𝐮𝐧𝐢𝐭𝐢𝐞𝐬 from incorporating AI into performance management:
1. AI can sift through heaps of information to collate performance trends incredibly quickly, dramatically reducing the administrative workload for managers and HR. AI can be trained to rapidly gather a holistic, summarised picture of an individual’s performance:
o their achievement against goals, objectives, metrics, OKRs and KPIs
o the feedback received from peers, direct reports, clients and bosses
o the comments written about them in check-ins and formal reviews
2. AI can then create an objective*, unbiased report based on a trawl of all this information. AI’s dispassionate lens could prove very helpful in alleviating a manager’s unconscious bias [*important caveat re. the ‘objectivity’ of AI below]
3. AI could increase an employee’s self-awareness by enabling them to understand their own performance and giving them comparative data in relation to peers. If an employee is able to digest this information in private before attending a performance review discussion it might allow them to be more attentive to their manager’s feedback, avoiding the ‘fight or flight’ tendency people tend to have in appraisals
𝐑𝐢𝐬𝐤𝐬 from using AI in performance management:
1. *Perhaps the biggest risk and misunderstanding is to believe that AI is fool-proof and ‘objective’ simply because it’s not human. It is important to understand that humans built the AI model, with all their foibles and biases. The AI model will simply emphasize and de-emphasize what it’s been taught to look for. No output is entirely ‘pure’; AI can be known to ‘riff on a theme’ and skew the importance or relevance of a particular subject
2. AI cannot solve the issue of ‘rubbish in, rubbish out’. Poor quality or minimal information for AI to draw from means any summary report will be equally poor quality
3. There is a risk that inept managers will take the AI-created initial summary of performance as the ‘gospel truth’ and finished product. Managers must understand that whilst AI can give them a straw model, it is 𝒕𝒉𝒆𝒊𝒓 job to critique, correct and humanize it ready for human consumption
To mitigate these risks, comprehensive education and training programs will be vital to increase the understanding of AI and its impact to open people’s eyes wide open to both its opportunities and threats. Change management is also critical to help people adapt to, and influence, these exciting new technologies.