AI in the Workplace: Productivity, Work Intensification, and Organisational Design

Artificial intelligence is no longer just an experimental tool to boost productivity. It is becoming core infrastructure inside organisations, changing how work is done, how roles are defined, how decisions are made, and how value is created.

Based on research from McKinsey & Company, Harvard Business Review, Boston Consulting Group (BCG), and Zylo, this article summarises current evidence on AI adoption, its impact on productivity, how it changes behaviour at work, the risks it introduces, and what organisations must do to govern it responsibly.

1. From Generative Tools to Agentic Systems

McKinsey’s “Superagency in the Workplace” describes a major shift in how AI is used at work. AI is moving from simple assistant tools to more autonomous systems that can complete multi-step tasks on their own.

In the past, AI mainly helped with tasks like summarising text or making recommendations. Newer systems go further: they can interact directly with customers and carry out actions themselves, including fraud checks, processing payments, or coordinating logistics.

1.1 Adoption Paradox: High Awareness, Uneven Usage

McKinsey’s survey data shows that most employees and executives are familiar with generative AI tools. However, leaders often underestimate how much employees are already using AI in their day-to-day work.

C-suite leaders estimate that only 4% of employees use generative AI for at least 30% of their daily tasks. In contrast, 13% of employees report doing so. Employees are using AI at roughly three times the rate leaders assume.

1.2 Macroeconomic Productivity Effects

McKinsey estimates that if generative AI is fully integrated into production systems, it could raise labour productivity in developed economies by around 15%.

This level of improvement would be comparable to previous general-purpose technologies, such as electricity or the internet. However, the transition may involve short-term disruption, including job displacement and adjustment costs as roles evolve.

2. The Behavioural Shift: AI and Work Intensification

Harvard Business Review argues that automation does not automatically reduce workload. Instead, it often changes the nature and pace of work.

2.1 Task Expansion

AI makes it easier to attempt tasks that were previously unfamiliar or outsourced. As a result, employees often take on additional responsibilities. Job scope expands informally, even when official role definitions remain unchanged.

2.2 Boundary Permeability

Because AI tools are fast and conversational, starting a task feels effortless. Small tasks begin to spill into breaks and after-hours time. What seems like quick interaction can accumulate into significant additional workload.

3. Risk Taxonomy

Zylo’s research on workplace AI identifies several operational and strategic risks:

• Data Exposure: Employees may unintentionally share sensitive information with external AI systems.
• Licensing and Software Waste: Overlapping subscriptions and unnecessary costs as teams independently adopt tools.
• Skill Atrophy: Heavy reliance on generative AI may weaken critical thinking and problem-solving skills.
• Bias Propagation: AI systems can reproduce and amplify biases present in their training data.

4. Organisational Design: Building an “AI Practice”

HBR suggests that organisations should develop an internal “AI practice”, a clear set of norms that defines how AI is used, at what pace, and under what oversight. The focus is not just productivity, but sustainable integration.

Three practical interventions are emphasised:

• Intentional pauses: Structured checkpoints to preserve reflection and quality control.
• Sequencing: Batching company notifications to reduce context switching.
• Human grounding: Maintaining structured human discussion to offset isolation.

5. Sectoral and Functional Use Cases

Common applications include knowledge retrieval, drafting communications, career planning, and automated reporting. In each case, AI acts as a cognitive amplifier, reducing the time between a question and a usable insight.

5.1 Workforce Recomposition

As AI becomes embedded, the skills that create comparative advantage shift toward strategic analysis, emotional intelligence, and interpersonal judgment. Work itself is not disappearing; it is being reorganised around human supervision of intelligent systems.

5.2 Strategic Implications for Organisations

For innovation-oriented institutions such as City Innovation Hub, the evidence suggests shifting from ad hoc experimentation to establishing structured routines, investing in AI literacy, and defining formal usage policies before scaling adoption.

Conclusion: The central issue is not whether AI will reshape work. It is whether organisations will consciously design that transformation, or allow speed and fragmentation to dictate outcomes.