AI’s causing a racquet at Wimbledon, but it’s a trend skirting the line in PR and business too

AI has been causing quite the ‘racquet’ this week in South West London. On Sunday, Wimbledon’s electronic line-calling (ELC) system – which uses AI and is replacing human line judges for the first time – failed to call ‘out’ for a ball that was well out. The decision, it was revealed, was due to the system being turned off by humans.

If you thought that was the end of it, in another match on Tuesday the AI ‘voice’ surprised everyone when it murmured ‘fault’ during a point being played. Technology was brought in to improve accuracy and, on the whole, it has. So, why is it being faulty? 

Zoe Kleinman penned a lot of good thoughts in her piece discussing why we don’t trust technology in sport. The tension arises when there is a lack of human control or input in AI-generated decisions. It taps into very valid insecurities infused into our collective consciousness about robots replacing humans. And ultimately, as many pieces have latched on to, it’s the human nature of imperfection and unpredictability that keeps sport engaging and exciting. 

But I want to hit on an area that’s missed the mark slightly in press coverage on the matter, before exploring how it relates to businesses using AI. 

The AI objective-subjective line 

Many of the articles I have seen compare tennis’ ELC system to football’s much-maligned video assistant referee (VAR). Understandably and correctly, the parallel comes down to implementation issues and how humans use and oversee the technology. In turn, for tennis, this has led to less surety from some players on its accuracy. 

But the ELC-VAR comparison is not wholly aligned. ELC actually falls more in line with goalline technology, a revelation for the football world. Like tennis, the ball-tracking cameras are operated by the firm Hawk-Eye – and they’re looking out for one thing: whether the whole of the ball crosses the goalline. It’s similar in tennis, just the other way around: any part of the ball has to be touching the line (albeit with more lines and variables). 

Out of thousands and thousands of matches, there’s only been one incident that comes to mind in football’s top flight – a very costly one, mind you – where the technology has failed. Apart from that, who knows how many controversies have been avoided by the near instant buzz the ref’s watch receives to say when a team has scored a goal. 

You can’t dispute the objective decision – and that’s the crux of the matter around AI. Objective tasks are where it thrives. When subjectivity is served into the mix, however, that’s when issues start to get sprayed across court. Public frustration froths up when a decision is so obviously wrong, yet the umpire or VAR doesn’t overrule it – why? It’s a predicament that is mirrored in the business and AI arena too. 

Defining AI implementation and messaging 

Regarding AI and business, a similar issue applies. If teams become too dependent on AI and technology, they don’t know how to act or respond when it misfires. What’s more, creativity, judgement, fairness, emotion, are all very human qualities. So when AI attempts tasks that are more subjective in nature, it can noticeably fall short. The goal is working out where AI can be an asset and where human intervention is crucial. It’s great in healthcare and diagnosing diseases, for example. Writing a blog like this, however, less so (so I hope), because it’s a human opinion. 

Bringing in a PR angle, there’s also an intrinsic issue with messaging. AI is a buzzword to generate interest in a brand, but can also be turned against it – so finding clarity in your messaging and how you promote yourself is incredibly important. Clickbait headlines tend to latch onto AI as the villain behind any mishaps or feed the brouhaha around the technology. But in the tennis example, for instance, AI only comprises a part of ELC, helping to conduct data analysis – the issues have all stemmed from human error or cameras being blocked. 

For both business use and PR, it’s worth noting just how broad a category AI is, encompassing machine learning, deep learning, generative AI models like ChatGPT and, more recently, AI agents. It’s both true that AI can generate highly accurate data analysis while, concurrently, it can produce wildly inaccurate information. Again, it’s the crossover between objectivity, subjectivity and, crucially, how the tools are used and defined. 

People want fairness and the right decisions. But they also want the liveness and humanity of sporting events. It’s the same with business. What this exemplifies is that AI has to enhance the humanity and value of what you do, not diminish it.