Does every company need a head of AI?

AI technology is becoming more visible within the workforce, with a big increase in job descriptions that mention artificial intelligence.

LinkedIn’s ‘Future of Work’ report states that the platform has seen a 21x increase in the share of global English-language job postings that mention new AI technologies such as GPT or ChatGPT since November 2022. More broadly, it notes, the number of AI-skilled members was 9x larger in June 2023 than in January 2016, globally.

AI leadership roles in particular are on the rise. LinkedIn states that, in the last five years, the number of US LinkedIn members in Head of AI roles has nearly tripled. We’ve already seen this type of hiring activity from big companies ranging from Amazon and PayPal to Coca-Cola and Netflix.

But what’s needed for AI leadership roles of this kind? And how might the trend have a wider impact on companies and their data or IT departments?

To gain some insight, I spoke to some industry experts working with AI at Dropbox, Intercom and Yext.

‘The opportunities presented by LLMs are too big to ignore’

Leadership roles such as ‘Chief of AI’ or ‘Head of AI’ are a “valuable need,” Fergal Reid, Director of Machine Learning at Intercom, told Econsultancy. In fact, he suggests “the right parallel here is a non-Internet business in the 1990s creating a new digital department, and a ‘Head of Digital’.”

“For many businesses, the value of unlocking the opportunities presented by LLMs are too big to ignore. We’re going to see a lot of change organisationally, in terms of skills and culture, for companies to take advantage of this new technology. Having a dedicated role to navigate these changes makes a lot of sense,” he stated.

The right parallel here is a non-Internet business in the 1990s creating a new digital department, and a ‘Head of Digital’.

– Fergal Reid, Intercom

Nico Beukes, Managing Director of Northern Europe for Yext, agrees that the rise of AI leadership roles “isn’t an opportunistic trend, but rather a needed position due to the significant impact and transformation global economies will undergo due to the direct effects of AI in the workplace.”

Crucially, however, he says that these roles should not merely be appointed in order to steer AI investment or oversee implementation from a product or strategy sense, but to take on the responsibility of equipping employees with the necessary skills to use AI in their own roles.

Beukes points to a recent study, published by Goldman Sachs, which revealed how AI could replace the equivalent of 300 million full-time jobs, also highlighting the level of concern from the general public regarding AI and the workforce. “Because of this, it is a great time for organisations to implement a head of AI to support employees in upskilling and learning how to use the technology to support their existing roles,” he explained.

[I]t is a great time for organisations to implement a head of AI to support employees in upskilling…

– Nico Beukes, Yext

“Let’s take the Industrial Revolution in the 19th century as an example. While manual jobs were replaced by machinery during this time, historically speaking, there was an increase in wages and more positions developed to operate those machines. This will also happen with AI, which is why businesses require AI leaders to improve employment opportunities while supporting corporations’ long-term growth goals.

“The critical objective for AI leaders is to remain at the forefront of their respective industries amidst the disruptive influence of AI or, at the very least, avoid being left behind.”

A balance of technical and change management skills is needed for the role

When it comes to the expertise needed to take on an AI leadership role, a “strong technical foundation” is essential, suggests Beukes.

“This includes having a deep understanding of machine learning and data analysis and experience working with various data management tools, alongside the awareness of AI regulations and ethical AI considerations.”

Again, however, Beukes reiterates the difference between a leader in AI and those implementing it. “It is the ability to foresee how the technology can create value for the organisation and the strategic steps or best course of action to upskill employees with the technology while supporting a company’s business objectives,” he said.

“Ultimately, those who can balance technical awareness with a strong acumen in the best business practices to apply them thoughtfully and innovatively are most likely to succeed as AI leaders.”

AI systems are imperfect, and a Head of AI needs to be able to measure their quality to make improvements and seek out opportunities to deploy them.

– Fergal Reid, Intercom

Intercom’s Fergal Reid says that qualitative skills are required first and foremost. “AI systems are imperfect, and a Head of AI needs to be able to measure their quality to make improvements and seek out opportunities to deploy them,” he said.

“Then, basic change management skills. The pace of technology innovation had slowed significantly in the 2010s, but recently has begun to speed up massively again. Where folks were used to dealing with stable technology, they now have to get comfortable with ambiguity and change as the pace picks back up.

“A major asset for navigating this change is an engineering mindset. With the technology changing so fast, there’s a lot of false information out there. Someone who will actually get their hands dirty and try to use modern AI will be a big advantage to the organisation.”

AI leaders should both “understand the technology’s current capabilities” and “envision future opportunities,” says Chris Noon, Global Head of GTM Intelligence, Data Science and Innovation at Dropbox EMEA.

But at the same time, he adds, “they need to be a strategic programme manager who can run a company-wide function that’s designed to identify pain points within the business that should be solved with AI.”

Organisations remain unsure where the role should sit, but collaboration is key

While many organisations might place AI roles within existing data or IT departments, the danger is that this could create limitations on how AI is perceived and implemented across the wider business.

Noon told Econsultancy that adding AI to the title of ‘Head of Data’, for example, “might not be optimal.”

“The Head of AI role should really stem from the business strategy team rather than the IT team,” he suggested. “The power of this new generation of AI lies in optimising time-consuming, expensive, and repetitive tasks. As AI learns how you do things, it can quickly start streamlining administrative tasks. But, at the same time, AI is not a panacea.”

In order to embrace AI most efficiently, Noon says that “businesses require a function that identifies, centralises, and prioritises the challenges that a company needs to address.”

The Head of AI role should really stem from the business strategy team rather than the IT team.

 

– Chris Noon, Dropbox EMEA

“If AI is the right solution, it will then need to collaborate with the right technical teams to determine whether existing AI capabilities can be purchased or realised using existing tools. Or, whether you need to build something from scratch.”

Yext’s Nico Beukes suggests that the Head of AI should sit “within an enterprise’s specific business unit or within a tech company’s product and innovation team – and report to the Chief Strategy Officer (CSO) to maximise value within a corporation.”

This way, he says, it ensures AI practices are fully aligned with strategic objectives as well as geared towards driving long-term growth. “The Head of AI is responsible for implementing the most effective AI approaches internally and externally and identifying opportunities to drive value and innovation across the organisation,” he states.

Success in AI leadership will stem from challenging paradigms and changing processes in the long-term

So, will AI leadership roles become commonplace in future, or will they become another niche tech trend, similar to the spike in Web3-related positions?

AI leadership could become a key part of a wider and long-term AI strategy for many organisations, but only if roles orchestrate real change. As Chris Noon from Dropbox puts it, “To make [AI] a success for organisations and people, it will be vital for these new Heads of AI to challenge paradigms and uncover AI’s potential to elevate processes to the next level.”

Intercom’s Fergal Reid suggests that the valuable and practical nature of AI technology means that employing a ‘Head of AI’ is much more valuable in comparison to other roles that are more about “speculative new things you might want to do.”

“AI helps you do the jobs you’re already doing, so organisational change and dedicated roles to support AI make sense,” he said.