For AI and automation to have a positive impact, transparency and psychological safety a must, says panel

Image:

l to r: Claire Taylor-Evans, Clare Schramm Fergus. Bisola Kayode, John Leonard

An expert panel at the recent IT Leaders Summit took audience questions on the impact of AI on jobs and the future of work.

In addition to technical expertise, the panel, chaired by Computing’s John Leonard, could also provide legal expertise in the form of Claire Taylor-Evans, an employment lawyer at Boyes Turner, who was able to provide insight on the increasingly vexed question of automated screening of CVs. Taylor-Evans thought that, in contrast with a lot of the concern being voiced by jobseekers, the potential here was positive.

“It’s obviously a labour-intensive process and AI doesn’t have fatigue and it’s not influenced by emotional decisions. Done properly, it can be effective in removing any unconscious bias, which I think is really exciting.”

‘Done properly’ is the key phrase here. Taylor-Evans continued:

“It is a high-risk activity because with sometimes very little human input it’s making fundamental decisions about people’s livelihoods, their legal rights. So it is really important that there are appropriate safeguards in place and that’s where I think that HR and business leaders really need to make sure they are.”

Important it certainly is because as Taylor-Evans acknowledged, bias has been an issue in the past, perhaps most famously in an algorithm Amazon used for hiring until the company realized that it was storing up some legal difficulty for itself by simply replicating its existing male dominated hiring patterns.

“I think it shows why monitoring, accountability and human oversight is so important,” she said. “A lot of these systems are developed outside the UK so it’s vital that developers are fully aware of the jurisdiction and law in the country that their product is being developed for.”

Bisola Kayode, Technology Risk, Cyber Risk Assistant Manager at Deloitte, outlined one of the risks associated with the prospect of AI taking away from traditional IT jobs.

“Maybe at some point when AI has developed, we can have the general workforce skilled to a point where a normal user could perform super user activities. Traditional IT tams would have privileged rights to key resources within their IT environment so if a normal user is able to access those resources, from a cyber security perspective that’s a threat because of the lack of training.”

Is AI really responsible for tech job losses?

The audience at ITLS also wanted the input of the panel on the wave of tech job losses that we’ve seen over the last 18 months. These job losses have often been attributed to AI, and in some cases (hello Salesforce) companies have courted publicity to exactly this effect. But are all these job losses attributable to AI or is it just a way to cut costs or present declining sales revenue as innovation?

Bisola Kayode said: “A lot of it is being driven by vendors putting more AI and automation into their applications, it’s not being driven by their customers. I would also say that for some of the big guys, they are not making cuts because of AI. In some cases, they are making them because of reduced profits. In consulting for example, there has been a reduction in sales revenue. So, companies lay people off and say it’s AI.”

Ultimately, technological change will only succeed if people remain at the heart of it.

Clare Schramm Fergus, MD, Banking Technology Platforms at Lloyds Banking Group said: “I think when we remove the barriers to collaboration, whether that’s by providing people with sandbox environments, getting people thinking n different ways, doing some training around AI, whatever it may be, these are the companies that are going to be creating meaningful step change. Enable your current workforce and allow them access to the tools that they need.”

Claire Taylor-Evans agreed: “I think there is a lot of fear out there and if you as a leader can make sure that you are bringing your staff with you and that you are transparent and can create a psychologically safe environment where they can challenge respectfully, I think that’s the best approach.”