The Future of AI is Human-Centred

AI value doesn’t come automatically

Artificial intelligence has made its way into boardrooms, but its value often remains unrealised.

Despite significant investment, 70-80% of enterprise AI projects fail to deliver expected outcomes. This isn’t a lack of capability, the problem is the approach.

Too often organisations chase automation for cost savings without considering the human experience.

This results in systems that may deliver quick efficiency gains, but:

• are difficult to trust,
• complex to adopt, and
• quickly fall out of step with real-world needs.

We’re past asking whether to use AI. The real question is how do you make it work in practice?

Designing AI that works for people

Human-Centred AI (HCAI) combines the power of automation with the insight, judgement, and context of the people who use it.

HCAI is about designing AI to maximise both automation benefits and human value simultaneously. It’s not about limiting technology, but about deploying it strategically for lasting impact.

AI is a force multiplier: it handles the routine, scales the repeatable, and clears the way for people to focus on what matters most.

HCAI builds on the strengths of both sides:

• ‍AI excels at speed, scale, and consistency‍
• Humans excel at creativity, complex problem-solving, and relationships.

Effective systems combine these strengths, so that AI removes friction and people can focus on high-value, human-centred work.

At CEC, we believe good design starts with people. HCAI is grounded in the human-centred design (HCD) methodology, which prioritises understanding users before determining what technology should be built.

HCD involves deep empathy, iterative prototyping, and continuous testing to ensure solutions meet human needs, not just technical possibilities. HCAI builds on this by designing AI systems that foster trust, transparency, and meaningful human-AI collaboration.

Three principles of Human-Centred AI

The following principles guide the design and deployment of AI to create systems that are trusted, adopted, and genuinely valuable in the real world.

Every practitioner working with AI should keep these principles front of mind, regardless of scale, progress or budget.

A 5-step process for HCAI success

You need more than just cutting-edge technology to succeed with Human-Centred AI. You need a thoughtful, structured approach to integrate AI into real-world human systems.

This five-step process ensures AI is designed and deployed in ways that drive both human and operational value.

Step 1: Start with human insight

Start with deep qualitative and quantitative insight. Not just user needs, but also organisational systems, goals, and contexts.

This approach identifies the real pain points, constraints, and opportunities where HCAI can create the most value.

Step 2: Define objectives and evaluate opportunities

Think about opportunities through two different lenses:

• Human outcomes: Trust, ease of use, confidence, wellbeing, reduced friction, empowerment
• Operational outcomes: Accuracy, efficiency, speed, cost savings, compliance, adoption rate

This anchors the initiative in measurable goals that align human-centred design with commercial and strategic priorities.

Step 3: Design for smart allocation of tasks and responsibilities

Allocate tasks based on strengths:

• ‍AI strengths: Repetitive, data-heavy, rule-based, scalable tasks‍
• Human strengths: Relationship-driven, ethically nuanced, strategic or creative judgment

This supports in avoiding over or under-automation, and designs balanced systems that are both smart and usable.

Step 4: Build in transparency for users and feedback loops

Make AI systems understandable and governable by the people who use and oversee them. This means:

• ‍Expose how decisions are made. Use plain language explanations to show what the system does and why.‍
• Enable user feedback. Allow users to flag issues, correct outputs, or suggest improvements directly in the interface.‍
• Support oversight and iteration. Give system administrators tools to audit decisions, monitor performance, and adjust models or rules as needed.

These practices help build trust and encourage continuous improvement, making systems safer, smarter, and more usable over time.

Step 5: Monitor, adapt, and scale based on outcomes

Deploy measurement systems that go beyond technical performance. Track:

• ‍Human indicators, like satisfaction, trust and escalation rates.‍
• Operational metrics, such as time savings, quality gains and Net Promoter Score (NPS) lift.
• ‍System learning indicators, such as data model accuracy and blind spot emergence.

Iterate frequently, not just post-launch. Regularly check in with teams, and measure data from multiple sources, both qualitative and quantitative.

This will create sustainable value over time and avoid performance plateaus.

Why Human-Centred AI wins

The business case for these principles is compelling:

• 91% of users believe that being able to explain how an AI decision is made is critical to building trust
• Organisations that incorporate trust-building measures into their AI efforts achieve a 6.6x greater return on benefits.

Human-Centred AI delivers better outcomes in three critical areas:

1. Faster adoption, lasting results: People trust and adopt AI they understand. This means organisations reach ROI faster and maintain gains longer, avoiding the plateau effect that kills many AI initiatives.
2. Iterative improvements over time: HCAI incorporates feedback loops, allowing design and performance to be iteratively improved over time. This user outcome feedback will enable organisations to adapt and iterate over time.
3. Sustainable competitive advantage: HCAI protects quality, compliance, and trust. Satisfied customers need less support. Confident employees perform better. Transparent systems earn regulatory approval. These advantages compound over time and are difficult to replicate.

These advantages explain why innovative organisations are shifting from cost-focused automation to human-centred AI strategies.

Looking to apply HCAI in your organisation?

Start with our Human-Centred AI Game Plan: a step-by-step guide to identifying high-impact opportunities and building systems people trust and use.