Most staff use AI. Few rely on it. Here’s why that gap matters

AI's value compounds with use.

---

A few years ago, artificial intelligence in the office meant a chatbot on the company website. Today it sits inside our e-mail, our spreadsheets and our code editors, often without much ceremony. The question for business leaders is no longer whether staff are using AI, but how well.

Mweb, which adopted a fully remote business model in 2025, wanted an honest answer from the people closest to it: the company's own employees. In September 2026, Mweb surveyed 69 staff, from customer-facing teams to developers and data specialists, about how they use AI, what it does for them and how they think it is changing work.

The results point to a workforce that has embraced AI faster than many organisations realise, but one where the biggest gains are concentrated among a small group of confident, everyday users.

Adoption is broad: 84% of respondents use AI at work in some form. Just over one in five (22%) use it every day, while 62% use it every now and again. Only 16% say they never use it.

That everyday presence is reflected in how people feel about it. A total of 20% of respondents say they can’t live without AI, and a further 62% say it somewhat helps with their daily tasks. More than half (54%) say it has improved their productivity, time management or quality of work, with another 26% undecided.

AI's value compounds with use.

---

Staff also described using AI for Excel formulas, brainstorming, newsletters, data analysis and tailoring canned customer responses to the brand voice. One developer uses it to scan logs and other data that would be very hard for a person to work through manually.

The common thread is time. Respondents repeatedly described AI as a way to get to a first draft faster, find the right tone for a difficult e-mail or get unstuck on a problem. As one put it, employees now spend less time producing drafts and more time reviewing, correcting and adding context to what AI produces.

The most striking finding is the gap between frequent and occasional users.

Among employees who use AI every day, 87% say it has improved their productivity, time management or quality of work, and 60% say they can’t live without it. Among occasional users, those figures drop to 56% and 12%.

In other words, AI's value compounds with use. Daily users have moved beyond grammar checks into complex tasks: writing code, analysing data, drafting documents and working through multi-step problems. Several described processes that once took hours now take minutes, and one said AI has enabled them to take on more challenging data and coding work than before.

For businesses, this is the real opportunity. Getting more people from occasional to confident daily use is likely to deliver far more value than simply rolling out more tools.

If AI is so useful, why do only a fifth of staff rely on it? The open-ended responses point to four barriers familiar to many South African businesses.

Confidence and training. Some employees simply aren’t sure what they are allowed to do with AI, or how to start. One respondent suggested that a clear policy, paired with practical online training on how AI fits into day-to-day work, would give staff the confidence to use it.

Trust, but verify. Experienced users were the most vocal about checking AI output. Several stressed that you need to know what you are looking for and check the results, describing AI as an iterative process rather than an oracle. One respondent warned against treating AI as the ultimate authority at the expense of critical thinking.

Job security. Some respondents worry that AI will replace people or be used to monitor performance. Others see it differently: AI handles repetitive tasks, freeing people to focus on customers, creative thinking and solving harder problems. Several noted that AI skills will soon be as fundamental as computer skills, and that continuous learning will be part of every role.

The survey suggests that the businesses getting the most out of AI will be those that invest in people as much as in platforms. Based on what Mweb's employees told the company, that means:

• A clear AI policy that sets out which tools are approved and what information can and can’t be shared.
• Practical, role-based training that shows staff how AI fits into their actual work, not just what it can do in theory.
• A culture of verification, where AI output is treated as a strong first draft that still needs human judgment.
• Honest conversations about the future of roles, so that AI is seen as a collaborator rather than a threat.

As a fully remote business, Mweb sees this every day. When the connection drops, access to the tools that now underpin a growing share of daily work drops too. Reliable, uncapped fibre is what lets employees use AI as often as they need to, wherever they work.

Perhaps the clearest message from employees is that AI works best when people stay in charge. The tools can handle the repetitive and the routine, but judgment, empathy and accountability remain human. As one respondent observed, the companies that thrive will be those that keep the human element at the centre of their AI initiatives.

For South African businesses, the opportunity is not just to adopt AI, but to help every employee use it with confidence. That starts with the right policies, the right skills and the right connection.

Note: The Mweb survey was conducted online among 69 Mweb employees between 1 and 9 September 2026. Percentages are rounded to the nearest whole number.