Most business leaders agree they could make better decisions if they had better data at their disposal; data on customer preferences, competitors’ pricing practices, transport disruptions in the supply chain – ideally, data on everything.

And, while it may be true that high-quality data, when used effectively, can produce insights that lead to better decision-making, many leaders are actually struggling to make sense of the volumes of data at their fingertips.

A survey of more than 1,000 C-suite leaders by Signal AI, for instance, found that the most common barrier to decision-making is “overwhelming amounts of data”, cited by 44% of respondents. 

New AI and ML techniques can streamline and accelerate data analytics processes that are central to producing the insights that inform data-driven decision-making.

For instance, Publicis Sapient, a digital transformation consultancy, has made AI and ML technologies an integral part of both its client offering and its internal operations. The company’s models can predict employee churn, help to manage revenue and pinpoint marketing and sales opportunities using a mix of a client’s own data, third-party data and publicly available data. 

Augmented decision-making

Nigel Vaz, chief executive at Publicis Sapient, believes the most successful organisations in the future will be those that base their leadership and decision-making on high-quality data. In fact, he attributes his own company’s own growth (19% in FY 2023) to its use of AI to “facilitate better informed and more data-driven conversations”.

Drawing on the guidance and recommendations of AI at executive-level meetings can also help businesses mitigate the potential negative effects of hierarchy in the decision-making process. This could enhance agile thinking when critical actions are needed to outmanoeuvre opponents.

According to Professor Chris Tucci, who teaches the ML and AI executive-education programme at Imperial College London, the technology has the potential to completely change how C-suites reach their decisions. He believes the underlying models will likely become more “creative” over the next few years. 

“AI systems could soon be used for brainstorming ideas on developing a new market or customer segment, new products and processes, and new business models. It might even become a sort of adjunct for human decision-makers, in addition to the efficiency arguments for their use,” says Tucci. 

But could AI actually change the make-up of the C-suite? There has already been an influx of chief data and analytics officers (CDAO) in recent years. Data from NewVantage shows that 77% of Fortune 500 companies now have a CDAO, compared with 12% in 2012. Some organisations are even creating C-suite roles for AI leaders.

The technology may also further democratise decision-making across the C-suite. Vaz says the kind of holistic data approach implied by augmented decision-making would produce better decisions at every level. “This is less about how these tools change the way any individual makes decisions, and more about how they can unlock potential across the organisation.”

The ability to prompt AI systems and make sense of their insights, would then be essential for C-suite leaders in a truly data-driven organisation. Tucci says: “Leaders don’t have to be a coder or know how to build their own AI programmes. But they do need to appreciate how these things work from a broad point of view and how these tools might benefit the organisation.”

Of course, there are risks in using AI to inform C-level decisions, which often directly impact an organisation’s success or failure.

John Hill is founder and CEO of Silico, an AI-powered platform that helps businesses to simulate decisions and processes. He says the most common barriers to AI-augment decision-making are a lack of interpretability, difficulty integrating AI with existing systems and concerns about data privacy and bias.

AI has the ability to drive efficiencies across many of HR’s core responsibilities, from recruitment to workforce management. And, given the growing expectations on the function, HR teams may welcome the assistance.

According to Dr Aaron Taylor, head of Arden University’s School of Human Resource Management, the Covid years catalysed a change in the responsibilities of HR. Thanks to the impact of shifting workforce trends, many HR leaders found themselves at the centre of C-level decisions, the outcomes of which determined success or failure.

Taylor argues: “The profession’s evolution over the past 25 years – from ‘pay and rations’ to the strategic role it plays today – has, quite possibly, been more radical than that of any other business function.”

Recruitment and hiring

Robert Symons, senior vice-president for EMEA at SmartRecruiters, says AI can assist with recruitment and hiring in a variety of ways. For instance, AI tools can help hiring managers to screen volumes of submissions in a fraction of the time that it would take a human.

“A recruiter spends about seven seconds on a CV,” Symons says. “They skim it for things like a competitor’s name before moving on to the next one because they have such a large quantity to go through.”

He continues: “AI can analyse career paths, tenure, skills and many more data points, which all help to prioritise candidates who more closely match the intent of the job description – it goes far beyond keyword recognition.”

AI can also enable smarter chatbots and virtual assistants, making companies more accessible to potential applicants. 

Chatbots can operate across a range of messaging platforms. They are multilingual and, crucially, always available. Analysis of a chatbot used by one of Symons’ clients showed that 70% of candidate questions were answered outside working hours.

Chatbots can be used to handle early-stage interactions with candidates – for example, answering frequently asked questions, providing information about the company and the application process – but some recruiters are even using AI to streamline the interview process.

Symons explains: “As a recruiter, one of my biggest bugbears was scheduling interviews across multiple diaries. But AI scheduling tools can promote available slots and even push messages out to the candidates, read the response and go ahead with self-scheduling.”

Here, the value of AI lies in the time saved. “It can free up time for higher-value interactions between candidates and hiring managers,” Symons says. “That’s where the technology has the greatest impact.”

What about the challenge of making a job vacancy known to the right demographic? Targeting features on sites such as LinkedIn have been around for some time, of course, but AI is greatly enhancing the capabilities of these tools.

For instance, by drawing on past recruiting data, AI can identify the most effective channels for recruiters to engage suitable candidates. This can even be as granular as automatically making small adjustments to listings to ensure they reach the right person.

Finally, AI can enable better candidate testing. Psychometric assessments and personality tests have been gaining prominence in application processes for several years, but the arrival of AI is enabling employers and recruiters to get even more information about a candidate’s strengths and character from such tests.

How will AI impact employee experience?

What about employee experience – factors such as onboarding, training and development, culture and engagement? There’s evidence to suggest that AI can be used both to make employees’ day-to-day work lives more fulfilling and to create better opportunities for career development.

We know already that AI can help to increase workplace efficiency and productivity, for instance, by automating repetitive and resource-intensive tasks. But the other side of this coin is that delegating such tasks to AI can free up workers’ time to focus on activities that add more value for the organisation and enable better strategic decision-making.

Consider the HR function, for example. A 2023 report by Microsoft suggests that the HR team could use AI as a co-pilot to automate internal communication while also making it more effective.

How would such co-piloting work in practice? Take the use of so-called writer’s block AI to improve communications between HR and the workforce, for instance. This technology uses relevant information about the company and its employees to personalise messages and deliver these in the appropriate tone.

Such personalised engagement is important, as Taylor suggests that modern HR teams must place greater emphasis on the human elements of the function, which includes understanding what “makes employees tick and how that aligns with the company’s overall strategy”.

Working alongside AI, the HR team can map out more ambitious possible career paths for people in the organisation. For instance, the technology might spot hidden potential in an employee who’s been flying under the radar and prompt the team to alert that individual to an appealing internal role that would suit their talents and offer them a valuable development opportunity.

AI can also aid employee recognition – a wellbeing-boosting intervention that can be as simple as thanking someone publicly, yet is lacking in many workplaces. LinkedIn has reported achieving a 96% retention rate among employees whose work was acknowledged at least four times a year. Some AI tools are able to prompt HR teams to recognise and celebrate the contributions (or life events) of employees or ask their line managers to do so.

Of course, whether or not AI will be a net positive for the workforce remains an open question. Uncertainty around how AI will impact the workforce is one of the greatest sources of trepidation among industry experts and managers looking to roll out the technology.

For example, while more than a quarter (28%) of UK employees believe AI could make their jobs easier, roughly a third (32%) fear the technology could render their roles redundant.

Moreover, concerns about biases are especially relevant in HR-related tasks, so many of which directly impact the people – human workers – in an organisation. If an AI system develops biases thanks to incomplete training data, it could negatively impact, for instance, the diversity in an organisation, the composition of its leadership team and so on. These problems are magnified when decisions made on the basis of AI-powered insights reinforce prejudices or stereotypes about protected characteristics such as gender and ethnicity, even if unintentionally.

While leaps in GenAI have wowed much of the business world since ChatGPT was made widely available in 2022, marketing professionals have been profiting from the power of AI and advanced data analytics for years. 

Nonetheless, GenAI presents huge potential value to marketers. The possibilities are verging on bewildering, according to Eric Gregoire, senior vice-president and global head of digital in the consumer health division at Bayer, the pharmaceutical company.

“There’s never been a better time to be a marketer, because there is so much you can do with this emerging technology,” he says. “But you may also get frozen by where to start.”

Kevin Iaquinto, CMO at CommerceHub, an e-tail software developer, believes that the use of AI and data-driven solutions has already improved B2B marketing significantly in three areas, enabling “hyper-personalisation at scale, better targeting and better creative [output]”.

His view aligns with the conclusions of 2023 research published by McKinsey, which found a significant increase in the use of GenAI for lead identification, resulting in better targeting and personalised outreach. 

Steve Reis, a senior partner at McKinsey and co-author of the research report, says that AI is helping to answer one of marketing’s fundamental questions: who is the customer? This is particularly challenging in the B2B space, because each key purchasing decision in a business is typically made by several stakeholders, few of whom will have exactly the same priorities. 

The growing capacity of data-gathering and advances in real-time analysis promise marketers a much better idea of who the main decision-makers are likely to be and what matters most to them. This should in turn enable them to personalise their marketing communications.

Moreover, AI and data analytics can be applied at various stages of the sales funnel – the popular marketing model that charts the progress of each potential customer from awareness (learning of a product’s existence) all the way down to action (buying it). 

As an example of a lower-funnel application – that is, the point at which someone is about to make a purchase – Bayer partnered with Google to analyse navigation data supplied by visitors to its own website. The aim was to identify high-value customers based on behavioural traits, such as the time they stayed on each page, to learn who among them were seriously considering a purchase. Such insights enabled Bayer’s marketing team to target those individuals with personalised messages, leading to a double-digit percentage growth in its sales conversion rate.

Gregoire notes it wasn’t long ago that marketers would base their campaigns on intuition or “what you think is best for the customer. Now you actually know what’s happening in real time. You can make smarter decisions, test, learn and transform.”

But he warns that making the best use of advanced tech solutions is no easy undertaking.

Tips on preparing an AI-ready team

The challenge here for many marketing teams lies in moving from recognising the value of AI and data-driven solutions to implementing them, according to Gregoire. In the first two years of its AI adoption, Bayer focused on building effective partnerships with tech companies and adding new skills to the team, incorporating AI gradually into the mix. 

Iaquinto believes that marketing teams must first develop the right skills to fill the growing need for “prompt engineers, data analysts and marketing automations”.

Many CMOs have found it hard to persuade every member of their marketing and sales teams of the benefits of AI – and Iaquinto reckons it’s almost impossible to realise its full potential without your department’s total support. No matter what great leads you provide, “you won’t see the growth in the pipeline if sales teams aren’t committed to using it”, he stresses.

Every step in adopting these solutions must be clearly communicated across the entire function. It’s a matter of building a common understanding and then setting clear objectives and measurable key performance indicators.

For the sales and marketing function to make the most of AI’s power, it must work more closely with IT and data specialists in the organisation, according to Gregoire. This should help encourage managers to adopt a “test and learn” approach when trying out new applications for the technology, he adds. 

This means accepting the inevitable risk that any given experiment will fail. 

“You can learn a lot from what didn’t work,” says Gregoire, who reports that unsuccessful experiments are actually celebrated at Bayer. 

This is not about getting jubilant when things don’t go as expected, of course. Rather, it’s about seeking insights from failures instead of brushing them under the carpet; explicitly acknowledging that testing new tech is not without its downsides; and actively encouraging people to engage in the sort of calculated risk-taking that fuels successful innovation.

Machine made: the use of AI-created content

Creative automation is the use of technology to mechanise various aspects of content generation. By combining elements of AI, design automation and data-driven decision-making, it streamlines the production of marketing material. In doing so, it eases one of the biggest challenges that marketers face: how to deliver a reliable supply of relevant (ideally, personalised) content across several channels to attract and retain the attention of consumers in an always-on digital economy.

Achieving the desired quality and quantity of output is hard enough in normal economic conditions, let alone at a time when ad budgets are squeezed and beleaguered consumers tire quickly of monotonous messaging. For instance, a survey of US consumers by software developer Celtra found that two-thirds of respondents regarded brand ads as repetitive and didn’t want to see the same material over and over. 

By removing the repetitive manual tasks involved in the production process, creative automation enables marketers to create assets in the timely and cost-effective way required to deliver higher-quality content on a global scale. That’s the view of Gareth Davies, CEO of creative agency Leagas Delaney.

“The reality of creative automation is that it lets you deliver a product of consistent quality,” he says. “It typically allows you to generate assets 10 times faster than you would by using conventional manual processes – and somewhere in the region of 80% cheaper. When you think that much of the financial benefit of working in an agency is driven by large-scale creative production, it’s clear that automation will reshape our industry in the coming years.”

Despite such predictions, the World Economic Forum revealed in its Future of Jobs 2023 report that the global uptake of AI and automation has not been as fast as it had initially expected. In the marketing sector specifically, this can be attributed to one key factor: a long-standing concern, held by many in the industry, that the adoption of cutting-edge tech could have a detrimental effect on creativity. 

This concern has heightened since the emergence of ChatGPT. For marketing campaigns, creativity and novelty are key to forming connections with consumers that last. It’s not surprising that the idea of being able to pump out content faster, with minimal human involvement, has been greeted with some scepticism.

Felipe Thomaz, associate professor of marketing at Oxford’s Saïd Business School, believes that the sceptics have little to worry about in this respect. He suggests that, by taking care of the aspects of content production that are simply about asset management, automation will enable – and, indeed, oblige – marketers to devote more attention to the creative side of their work. 

Thomaz explains that GenAI and creative automation have significantly reduced production costs, meaning the marketplace is being filled with “high-quality noise”. He believes that, in this environment, truly creative ideas – the ones only humans can generate – will become increasingly valuable.

AI in market research

Market-research companies have found AI to be a useful analytical tool, particularly its ability to understand what consumers write on questionnaires and say in audio or video interviews. The technology can also reliably interpret their answers to reveal hidden insights. It can even suggest next steps.

Now, market researchers are testing AI’s ability to use synthetic responses of its own devising, effectively cutting human interviewees out of the equation. If their experiments prove successful, AI could provide near-instant low-cost ‘consumer’ insights, reducing the need to conduct costly surveys and, potentially, enabling brands to reach lucrative niche markets. 

To produce reliable responses, the technology must be able to understand the views of the target audience and provide results that match those elicited by traditional consumer research methods. The natural question at this stage of development is: can the synthetic data it will produce be trusted?

Market researchers at Kantar have taken the first steps in answering this. They prepared a set of questions and compared real data drawn from human surveys with responses given by OpenAI’s GPT-4 large language model (LLM). The queries they used covered a wide range of matters, such as whether the price of luxury holidays is off-putting and whether a given piece of technology helps the owner to connect with people who share their interests.

When asked about more practical issues, GPT-4 gave similar answers to those provided by the human respondents. But the more nuanced questions, requiring greater emotional reflection, produced significant differences.

Such results were what you might intuitively expect, notes Jon Puleston, vice-president of innovation at Kantar’s profiles division. AI is good for some parts of market research, but it’s limited if asked to adopt the persona of diverse human audiences. 

“It’s clear that there are risks to relying solely on synthetic data if you’re making a business decision that’s worth billions,” he says. “Real human insights still form the heart of good market research. A more realistic use case for synthetic data is as a tool to complement, rather than replace, traditional research – for instance, by boosting sample sizes in surveys, particularly for niche audiences.”

The experiments’ results so far indicate that the LLM’s outputs are only as good as the human-profiling data fed into it, notes Marius Claudy, associate professor of marketing at University College Dublin, who has been researching the impact of training on AI outcomes. 

Plus, while the technology can provide a good analysis of qualitative research, such as understanding what someone has said or written, it’s less effective at understanding the emotions that underpin people’s responses. This leaves the notion that AI could ever make traditional market research obsolete open to question. 

But concerns are not limited to whether future models can replicate real human responses. There are also legal considerations, warns Ben Travers, a partner specialising in IT matters at law firm Knights. While he shares researchers’ worries that AI bias may lead to poor outcomes, he is also troubled by the use of personal data found on the internet to build profiles.

“Businesses will need to ensure they have a clear legal basis for uploading any personally identifiable data to AI tools,” he says. “And all AI users must be alert to copyright issues. These apply to both the content fed into an AI and the content it produces. Just because this material is easily accessible does not mean that it’s lawful to copy it. Such content is not ‘fair game’ – copyright will enable the rights owner to control how it is used and disseminated.”

Applications in the sales function

There are two main types of AI typically used by sales teams: generative and predictive. 

GenAI helps with things like summarising calls and extracting information, such as agreed action items, and can automatically write personalised follow-up emails. It can also present key insights from calls, including competitor mentions, objection handling and pain points to help salespeople improve with every conversation. According to Eilon Reshef, co-founder and chief procurement officer at Gong, a revenue-intelligence vendor based in San Francisco, using AI to surface key highlights from previous calls can save up to 80% of sales reps’ call prep time. 

Predictive AI, meanwhile, does things like sales forecasting, taking human-led bias away from revenue predictions and expected deal outcomes. “AI-powered forecasting can ingest many unique buyer signals – including buyer conversations – to provide a much more accurate projection of deal outcomes,” says Reshef. He claims that Gong’s predictive AI solution uses more than 300 unique signals to predict deal outcomes with 20% greater precision than algorithms based solely on CRM data.

But the applications don’t stop there: AI can help deliver bespoke presentations to clients; cut down on research time; analyse and score leads based on various criteria to help prioritise prospective clients; and create dynamically adjusting pricing based on factors like demand and market conditions.

Then there’s audience intelligence. “AI tools can analyse your pipeline and the typical sales cycle at prospects to determine the best time to approach key contacts. It will also identify those most likely to convert,” says Claire Simpson, head of growth at tech-focused PR and marketing firm Hard Numbers.

As the market matures, more niche tech has come to the fore, such as Evolve from translation company RWS, which recently launched in beta mode. Evolve, whose early beta customers include Dell, is a linguistic tool that uses secure neural machine translation, linguist-verified quality estimation and large language models to translate complex languages with accuracy in near real-time. The tech is said to be the first-ever AI capable of translating languages within specialist fields like medicine or law at local-speaker fluency. This enables sales teams dealing with international audiences to be more efficient, according to Thomas Labarthe, president of RWS’s Language and Content Technology division.

But just because salespeople have more information or insights at hand, it doesn’t necessarily mean they’ll be better at establishing trust with a buyer or having the personal skills required to secure a deal.

“Sales is, after all, a contact sport. And the decision-making process remains deeply personal. Conversion can come down to chemistry, credibility, trust – any number of human factors. In this sense, AI-enabled sales will always need humans to put its insights into action,” Simpson says.

With more automation, these soft skills will become even more vital to the job. Martin Roe, chief executive of business process outsourcing provider CCI Global, says it is “alarming” that so many companies deploy sales tech without first making sure their sales staff have adequate training. Employee training therefore is essential to realising AI’s benefits in the sales function.

High-quality customer interactions

The use of AI in customer-service chatbots is perhaps one of the most common applications of the technology. Indeed, using AI for better customer interactions is considered by many to be a quick win – not necessarily groundbreaking, but a win nonetheless.

But when deployed with care, AI-powered customer-service tools can create considerable value for organisations. According to Eric Jorgensen, vice-president of enterprise sales, EMEA, at Zendesk: “The overarching trend is simplification and automation, leading to better-quality experiences and reduced costs.”

The technology’s ability to quickly analyse vast quantities of data and draw real-time insights is helping to make customer interactions more personalised and dynamic. This has enabled some firms to turn regular chatbots into digital agents that can cross-sell products and services and direct customer queries based on initial contacts in real time.

What’s more, about two-thirds (68%) of UK customer-experience (CX) leaders believe chatbots can help brands to build an emotional connection with their customers. For instance, AI can recognise keywords and adjust the tone of its communications with a customer to fit the context of the conversation. If the customer insists on speaking to a human, the AI can provide a summary of the interaction to a human agent, enabling a seamless experience.

Jorgensen stresses, however, that transparency is essential with AI-enhanced customer service. To maintain trust and provide a truly elevated experience, customers must be made aware that they are interacting with a digital agent.

Naturally, each customer query an AI can answer is one fewer that a human must handle. But this doesn’t necessarily mean AI will replace human agents. Jorgensen says: “[Human] agents will deal with high-complexity cases and digital agents will deal with high-volume cases. AI and related technology will help us to provide quick answers to less complex queries, leaving agents with more time to deal with complex problems, as well as upselling and cross-selling.”

The chief financial officer role is undergoing a dramatic expansion and the adoption of AI, encompassing everything from chatbot assistants to fraud detection, looks to play a critical role in that shift.

The technology has the potential to transform the finance function by automating rudimentary tasks such as transaction processing and auditing, thus enabling finance leaders to spend more time on tasks that create value. A 2023 Gartner survey predicted that 50% of organisations will use AI to replace “time-consuming bottom-up forecasting approaches” by 2028.

Research from Blackline, a software firm, found that more than a third (38%) of CFOs believe AI will enable better analysis of historical financial data, which will help to improve forecasting capabilities, and 35% think it will enhance audit capabilities by analysing vast amounts of data to identify patterns or potential errors. 

Emma Brown, CFO at Medius, a spend-management solutions provider, says AI has enabled her to handle cumbersome processes, such as financial close, consolidation, invoice-to-cash and intercompany reconciliation, more effectively. AI has also helped Brown to develop multiple scenarios for capital allocation as it provides strong visibility of where money is being spent in the business.

This is important as nearly 40% of finance chiefs admit they do not trust the accuracy of their firm’s financial data, and 98% do not have complete visibility of their organisations’ cash flow, according to another survey by Blackline. The findings show that data often comes from too many sources, meaning finance leaders cannot be certain that it’s all accounted for. Plus, there is still an over-reliance on outdated data-management processes in the function.

Kristina Majauskaitė-Adomavičienė, CFO at Paystrax, a fintech company, agrees AI can create opportunities to run the finance function better. She believes the enhanced auditing capabilities of AI-powered forecasting tools can help finance leaders to unlock more creative ways to solve problems, develop new skills and grow professionally. 

On top of that, at a time when more than a third of CFOs are concerned about staff turnover, Majauskaitė-Adomavičienė is hopeful that AI can improve employee retention and engagement across the finance team by enabling people to undertake more interesting and meaningful work.

There are plenty of AI tools designed to enable people to spend fewer hours on tedious tasks and more time on higher-value work. Take Microsoft’s Copilot. “A couple of thousand people on a financial planning and analysis team each spend one or two hours doing reconciliation each week. With the new Copilot, that takes more like 10 or 20 minutes per week,” according to Cory Hrncirik, Microsoft’s modern finance lead.

Barriers to adoption

Although many finance chiefs are excited about the prospect of an AI-enhanced finance function, they are also hesitant to place too much trust in AI tools, especially when it comes to decisions that could directly impact the integrity of the organisation’s books.

“CFOs want to see the process at play,” says Karim Ben-Jaafar, senior vice-president at software company Quadient. “They don’t just want the conclusions, they want to understand how the system arrived at the answer and they want to be able to challenge the inputs and assumptions.”

AI implementation therefore is being delayed at least in part by an inability to fully understand how an AI arrives at its decisions – the so-called black-box effect.

Moreover, finance leaders are facing significant digital-skills shortages in their teams. For instance, by 2027, CFOs expect half of their staff to be able to create and modify finance technology capabilities, according to Gartner Finance. But, at most organisations, less than 20% of finance workers currently have the ability to do so. 

Many CFOs are also responsible for ensuring compliance with constantly evolving regulations and ethical framework.

Some of these challenges cannot be overcome by finance leaders alone – for instance, the AI skills gap creates difficulties across the organisation and must be addressed holistically. Others, such as the black-box effect, are fundamental challenges for which there are no obvious solutions.

AI will be essential to enabling greater supply chain transparency, which itself is essential for mitigating disruptions and ensuring regulatory compliance, for instance. That’s according to Anthony Plummer, group chief technology officer at Ligentia, a provider of global supply chain solutions.

The technology can simulate countless what-if scenarios and assess their impacts on the supply chain. One way to do this is by combining AI and digital twin technologies. These are detailed simulations of real-world objects, systems or processes built with real-time data. They provide snapshots that can help firms to monitor threats, test different scenarios and improve decision-making.

So far digital twins have mainly been trialled in manufacturing, but companies such as SAP and Oracle are starting to explore their use in supply chains. The hope is that they can help to identify trade bottlenecks, predict fluctuations in demand and tackle transport and inventory issues.

Joseph Buckley, director at Control Risks, a global risk consultancy, says the potential benefits of using the technology to manage supply chain risk are “vast and verging on revolutionary”.

Buckley explains: “By combining critical data, intelligence and indicators from the technologies represented by digital twins, decision-makers will be able to make more effective, proactive and well-informed decisions.”

Consider supply chains in the shipping industry, Buckley says. Digital twin technology could help the maritime industry map supply chain vulnerabilities more effectively, prevent mechanical failures before they happen and identify optimal shipping routes using data from sources such as GPS, ports, warehouses and shore-side operations.

However, even the best AI tools can still be caught out by black-swan events – those that are rare, difficult to predict and have a wide impact. Because these events do not follow past trends or patterns, AI trained on historical data is unlikely to foresee them. But even when AI doesn’t understand what has caused a ship to suddenly change course, for example, it may still be able to predict the impact on the supply chain.

“AI now has data on, for example, what happens when the Suez Canal is shut and transit time increases by two weeks,” explains Plummer. Companies can use that data to build a basic understanding of how future disruptions will impact logistics.

He adds that AI-powered tools can also assist with demand forecasting, reducing the risk of stockouts or surpluses, as well as routine tasks such as tracking shipments and scheduling deliveries. 

But human expertise is still needed for complex decision-making, Plummer says. This will remain true even as the adoption of more advanced tools such as agentic AI systems becomes more widespread.

There are of course several challenges to implementing AI in supply chain operations. For one, many of the technologies that are expected to work alongside AI to enable complete supply chain transparency – blockchain, IoT and connected devices and data analytics – require further development before they can be deployed effectively.

Supply chains are also prime targets for cybercriminals, meaning that supply chains fuelled by potentially sensitive data carry significant cybersecurity risks. The threat of a digital breach is magnified by the integration of AI, which is both uniquely vulnerable to and enabling of cyber attacks, as we’ll see in the next section.