Why emotional intelligence matters more as AI reshapes the workplace
Artificial intelligence can analyse data, draft reports and answer customers within seconds. But as businesses automate more tasks, the ability to communicate, build trust and exercise human judgement is becoming a more valuable, if difficult to measure workplace capability.
A bank’s artificial-intelligence assistant can identify a missed loan payment almost instantly. It can calculate the outstanding amount, issue a reminder and recommend the next action.
What it may not know is that the customer has just lost a job, suffered a bereavement or received a serious medical diagnosis.
Deciding how to handle that conversation still requires more than data. It requires judgement, empathy and an understanding of how words are likely to affect another person.
Sources
Ghana Statistical Service: Ghana’s youth population and unemployment
University of Ghana: Emotional intelligence in Ghana’s banking sector
That distinction is becoming increasingly important as businesses use artificial intelligence to draft documents, analyse information, schedule work, respond to customers and support decisions.
Technical skills remain essential, particularly for employees responsible for building, operating or supervising AI systems. But employers are also placing greater value on capabilities that help people manage themselves, understand others and navigate difficult situations.
These abilities are often grouped under one broad and sometimes contested term: emotional intelligence.
More than being friendly
Emotional intelligence generally refers to the capacity to recognise and understand emotions, regulate one’s reactions and respond effectively to other people.
In practice, it can involve listening carefully, staying composed under pressure, delivering difficult information sensitively, recognising when a colleague is struggling and resolving disagreements without making them worse.
It is not the same as being friendly, charismatic or extroverted. A quiet employee may demonstrate strong emotional intelligence by listening, considering different perspectives and responding thoughtfully.
Nor is it a substitute for professional competence. An empathetic surgeon still needs clinical expertise. A considerate accountant must still understand financial reporting, while a manager who communicates well must also be capable of making sound decisions.
The argument gaining attention in business is that technical and emotional capabilities increasingly need to work together.
AI changes tasks more often than it removes entire jobs
Fears that generative AI will eliminate whole categories of employment have dominated discussions about the technology.
Research from the International Labour Organization offers a more measured picture. Its 2025 assessment found that one in four jobs worldwide had some exposure to generative AI, but said transformation not outright replacement was the most likely outcome because most occupations still contained tasks requiring human involvement. Clerical work had the greatest exposure.
That transformation could alter what makes an employee valuable.
If software produces the first draft of a report, the worker’s contribution may shift towards checking its accuracy, identifying bias, deciding what should be communicated and adapting the information for a particular audience.
If a chatbot answers routine questions, human customer-service staff may increasingly handle the complaints, disputes and unusual circumstances that automation cannot resolve.
In healthcare, an AI system may help identify patterns in medical information.
But a trained professional must explain the findings, respond to a patient’s fears and accept responsibility for decisions.
The technology may therefore reduce some routine tasks while increasing the importance of judgement, trust and communication in the work that remains.
Employers want technical and human capabilities
The World Economic Forum’s Future of Jobs Report 2025 placed AI and big data at the top of the fastest-growing skills.
But the report based on responses from more than 1,000 employers representing millions of workers also identified resilience, flexibility, agility, leadership and social influence among the capabilities increasing in importance.
This does not prove that every employer now treats emotional intelligence as a decisive recruitment requirement. Neither does it show that such abilities matter more than technical knowledge in every occupation.
It does indicate that companies expect workers to combine digital competence with capabilities needed to lead people, adapt to change and collaborate.
The World Economic Forum estimated that 22% of jobs could be disrupted by 2030, with 170 million roles created and 92 million displaced. These are employer-based projections rather than recorded outcomes, and their global scope means they should not be treated as a forecast specifically for Ghana.
Why this matters in Ghana
Ghana launched its National Artificial Intelligence Strategy in April 2026, with workforce readiness, skills development and responsible AI adoption among its priorities.
The government says it wants the country to become an African centre for AI development rather than remain only a consumer of technologies created elsewhere. Its programmes include AI training for public officials and the One Million Coders initiative.
The strategy comes at a time when young Ghanaians face significant pressure in the labour market.
Ghana Statistical Service figures indicate that people aged 15 to 35 constitute 36.9% of the population. Youth unemployment averaged 21.9% during the period covered by its 2026 release, compared with a national average of 12.8%.
Digital-skills programmes could help more people qualify for emerging work. But coding, data analysis and prompt-writing alone may not be enough to make them effective employees or entrepreneurs.
A Ghanaian software developer may need to explain a technical problem to a client who does not understand the technology. A journalist using AI must decide when an automated answer is unreliable or unfair. A banker must balance risk rules with the circumstances of a distressed customer.
Teachers and health professionals face even more sensitive decisions because errors or poorly communicated information can have serious consequences.
For managers, emotional intelligence may be particularly important when introducing AI into an organisation. Employees who fear that automation could eliminate their jobs are unlikely to be reassured by a technical presentation alone. They need honest explanations about what is changing, which roles will be affected and what support or training will be available.
Evidence from Ghana but not yet about AI
Research involving Ghanaian workers suggests that emotional intelligence was relevant in workplaces before the current wave of generative AI.
A peer-reviewed study by University of Ghana Business School researchers Kwasi Dartey-Baah and Benjamin Mekpor examined 234 employees in Ghana’s banking industry.
It found that leaders’ emotional intelligence positively predicted voluntary, helpful workplace behaviour and was negatively associated with counterproductive behaviour. The research was cross-sectional, however, meaning it identified relationships rather than proving that emotional intelligence directly caused the outcomes.
Another Ghanaian banking study involving 482 participants examined whether emotional intelligence affected how employees responded to incivility from colleagues.
There is also Ghanaian research involving nurses, midwives and managers at Korle Bu Teaching Hospital. However, much of the local evidence relies on questionnaires and studies of particular institutions, limiting how confidently the findings can be generalised across the economy.
Crucially, there is little publicly available Ghana-specific research showing whether employers have increased their demand for emotional intelligence because of AI adoption.
Claims that it has “emerged” as a leading capability in Ghana should therefore be presented as a developing business trend supported by international labour evidence not as a conclusively measured national shift.
Can emotional intelligence be taught?
People may differ naturally in how well they recognise and manage emotions, but research suggests that at least some related capabilities can improve through training.
A 2024 systematic review and meta-analysis found that workplace interventions could strengthen emotional competencies across different professions.
The researchers also warned that the studies varied considerably and were often of low methodological quality. They called for more rigorous randomised trials before strong conclusions could be drawn.
Effective development is likely to involve more than a motivational speech or a one-day workshop.
Employees may benefit from repeated practice in active listening, receiving feedback, managing conflict, recognising emotional triggers and communicating under pressure. Managers can reinforce those behaviours through coaching and performance reviews.
Organisations must then determine whether training has produced lasting changes in how people work not merely higher scores immediately after a course.
The problem of measurement
Emotional intelligence is more difficult to assess than whether someone can operate software or analyse a spreadsheet.
Some assessments ask people to rate their own empathy, self-control and social awareness. But individuals may exaggerate their strengths or lack an accurate view of how they behave.
Other tests ask candidates to identify emotions from faces or decide how a person might respond in a hypothetical situation.
A systematic review of emotional-intelligence measures found significant differences in how the concept is defined and tested. The OECD has also cautioned that self-assessments of social and emotional skills have limitations involving comparability, validity and interpretation.
This creates risks when employers use emotional-intelligence scores to hire or promote workers.
Cultural background can influence how people express emotions, communicate disagreement or show respect. Assessments could also disadvantage neurodivergent candidates or people whose communication styles differ from those favoured by test designers.
A carefully structured behavioural interview, references and evidence from actual work may provide a more useful assessment than a single commercial test.
AI can imitate empathy but is that enough?
The boundary between human and machine capabilities is not fixed.
AI systems can already identify emotional language, produce sympathetic responses and adjust their tone. Some research has found that large language models can perform strongly on tests designed to measure aspects of emotional understanding.
But producing an empathetic sentence is not the same as experiencing concern, understanding the full social context or accepting responsibility for the consequences.
An AI assistant may generate an appropriate apology. A human employee still needs to decide whether the apology is sufficient, whether an exception to policy is justified and how to repair the relationship.
The business advantage may therefore lie not in humans trying to outperform AI at every task, but in knowing where automation should end and human accountability should begin.
When “emotional intelligence” becomes emotional labour
The enthusiasm surrounding emotional intelligence also deserves scrutiny.
Employers could misuse the term to demand that workers remain cheerful and accommodating in unhealthy conditions. Customer-service employees may be expected to absorb insults, calm angry clients and repair problems created by understaffing or faulty automated systems.
Workers who question decisions could be labelled emotionally unintelligent simply because they are direct or unwilling to conceal disagreement.
The pressure may fall particularly heavily on women and lower-paid service workers, who are often expected to perform more caring and relationship-management work.
Emotional intelligence should not become a way for employers to shift responsibility for poor working conditions onto employees.
The ILO says excessive demands, limited control, inadequate support and unfair treatment are identifiable workplace hazards that organisations can change.
A business cannot compensate for a harmful culture by teaching workers breathing exercises and asking them to communicate more positively.
The employee AI cannot replace may still use AI
The emerging workplace is unlikely to offer a simple choice between technical intelligence and emotional intelligence.
An employee who understands people but cannot use modern tools may struggle. A technically accomplished worker who cannot collaborate, explain decisions or recognise the human consequences of those decisions may also face limitations.
For Ghanaian businesses, the challenge is to develop both.
Workers will need sufficient AI literacy to understand what automated systems can do, where they fail and how to check their outputs. They will also need the confidence and judgement to challenge those systems when the result is inaccurate, unfair or inappropriate.
That combination, technical capability supported by human judgement may be more valuable than either skill set alone.
AI can draft the message, analyse the customer’s history and recommend a response. It is still the human being who must decide whether sending it is the right thing to do.