16/09/2025
Summer is winding down. The leaves are slowly turning. You know what that means: it’s “back to school” season! And what better time than now to look at how learning and knowledge are changing thanks to AI?
In the workplace, companies have been struggling for years to overcome shortfalls in skills and knowledge. Fast-changing industries like banking, insurance, manufacturing and retail have been among the hardest hit. We’ve seen chronic shortages in key areas like IT, cybersecurity, compliance, claims adjusting, engineering and data analysis.
It’s no wonder CIOs at banks, insurers, manufacturers and retailers are pouring budget into AI initiatives to bridge the gap. But can AI live up to expectations in the short term? And what do organisations need to do to ensure their AI investments pay off going forward?
In this first instalment of our series on the 5 actual truths about AI in the workplace, we’re taking an in-depth, realistic look at how to really multiply skills in the workplace with AI.
The talent deficit is here to stay
Skill shortages are slowing growth in banking, insurance, manufacturing and retail – dynamic sectors that rely on an adaptable, well-trained workforce.
The talent deficit shows no signs of ending anytime soon. According to Gartner’s 2025 AI Hype Cycle report, by 2030, half of enterprises will face irreversible shortages in at least two critical roles due to GenAI-related skill erosion and uncompetitive pay. CIOs across all four industries already say that IT and data talent gaps are a top barrier to delivering on business expectations.
Even in areas like GenAI and agentic AI where budgets are steadily growing, leaders are struggling to close the gap. Insurance CIOs say skills shortages are one of the biggest obstacles to scaling AI projects, while retail CIOs say frontline enablement is being slowed by the lack of staff who can interpret or trust AI outputs, which is why the need to multiply skills in the workplace is more important than ever.
Organisations increasingly find themselves in a tough position: on the one hand, they want to quickly adopt AI to multiply business-critical skills within their workforce. On the other, skills shortages make AI adoption less effective. Let’s look at how leaders in each industry can break the vicious cycle and actually put AI to work.
Banking: Reinforcing compliance and risk management
Faced with ongoing skills shortages, 92% of banks are now investing in technologies such as generative AI and AI agents to strengthen compliance and risk management, according to Gartner’s 2025 survey of Finance sector CIOs. So far, only one third of CIOs say their initiatives are living up to their CEO’s expectations.
This raises the question: what are those CIOs getting right that everyone else is getting wrong? For one thing, they’re focusing on the human side of AI adoption.
The most effective AI initiatives in banking don’t set out to replace humans or devalue their work. On the contrary, they position human decision-making as the gold standard. AI and machine learning can speed up fraud detection and reduce false positives, but these technologies should serve as a filter, not a judge. The real value comes when AI detects and escalates the cases worth attention, and trained staff make the final call.
Insurance: Moving beyond automation
Half of insurers have already rolled out AI or GenAI tools, and nearly 90% plan to boost budgets again in 2025, by an average of 38%, according to Gartner’s 2025 survey of Insurance CIOs. For example, Aviva has deployed over 80 AI models, slashing its liability assessment time for complex cases by 23 days on average, whilst improving routing accuracy by 30%. As a result, they’ve seen a 65% reduction in customer complaints.
Automation clearly accelerates key processes, but in an industry where trust is everything, working faster doesn’t always mean working better. As use cases for AI become more mature, forward-thinking CIOs are focusing on potential downsides like bias, accuracy drift and weak explainability.
To mitigate risk, insurers must put bias checks in place, set clear rules for when humans step in and keep proper records so regulators and customers can see exactly how each decision is made. AI upskilling should also be a key focus: to achieve its impressive results with AI, Aviva invested in over 40,000 hours of employee training.
Manufacturing: Introducing super-human efficiency
Manufacturers are struggling with chronic shortages in engineering and other key roles. No surprise then that 83% of Manufacturing CIOs are investing in AI, according to Gartner, in areas like product cycle optimisation, automated compliance reporting and quality monitoring.
Predictive maintenance has become a go-to use case, with major manufacturers like Agilent reporting up to a 51% reduction in downtime as a result. AI picks up on patterns people would miss, flagging faults before they cause breakdowns. For plant managers, that means fewer unexpected stoppages, which frees up crews to focus on higher-value work instead of constant troubleshooting.
While super-human efficiency sounds like a big step forward, manufacturers know better than to put too much faith in a single tool. Some are concerned that over-reliance on AI will erode core expertise over time. To make sure they always have a backup plan in place, they’re investing in multiply skills in the workplace to make sure they still get hands-on experience, even as they increase investment in AI-powered maintenance.
Retail: Building a fully AI-enabled workforce
Retail is one of the world’s most data-driven industries and has pioneered many AI use cases, from dynamic pricing and supply chain forecasting to e-commerce personalisation and market analytics. Amazon considers AI so crucial that it has launched a global project to educate 2 million people with critical, future-proof AI competencies.
Within the company, Amazon Web Services (AWS) takes a simple approach to AI upskilling:
- Everyone benefits from GenAI
- Prompt engineering is a must-have skill
- Use social media as an education platform
- No new tools without new training
While these rules sound easy enough, surprisingly few companies get them right. Gartner reports that 69% of CIOs believe employees are using unauthorised AI tools on the job – a clear sign that official AI policies and frameworks are lacking. Organisations in retail and beyond can all learn from AWS’s inclusive, employee-focused approach to AI adoption and change management.
Facing the truth
Most companies are still trying to figure out how to achieve value with AI. The ones who have already made it understand that AI is about amplifying people, to multiply skills in the workplace. That means embedding AI into the workflows employees already use, and just as importantly, providing targeted education and change management to create the cultural conditions for AI to thrive.
Frequently Asked Questions (FAQ)
It means using artificial intelligence tools and systems not to replace human skills but to enhance, accelerate and scale what employees can deliver, from learning new competencies, improving decision-making, to automating repetitive tasks, so people can focus on higher-value work.
Many industries, banking, insurance, manufacturing, retail, are facing chronic shortages in critical roles such as IT, cybersecurity, data analysis and more. AI helps bridge the gap by enabling better efficiency, speeding up processes, and allowing staff to learn or apply skills in novel ways.
Risks include over-reliance on automation, decision-making without sufficient human oversight, issues with bias and accuracy, and a loss of core expertise if employees are not kept engaged or trained properly. Ensuring clear governance, human judgement, transparency and ongoing education helps mitigate these risks.
