Accelerating human-centred AI: Transforming through trust

In our recent AI survey of 300 business leaders, 87 percent told us they have concerns around AI ethics and adoption, 83 percent have an unclear strategy of how to integrate AI, and 80 percent fear pushback from employees. These concerns are understandable. The worries about AI are writ large in its very name – that it's artificial. Yet AI needn't be at the expense of humans.

We explored this topic in our recent webinar, ‘Human-centred AI: How can we take the artificial out of artificial intelligence?’ In the session, we delved into the essence of human-centred AI, exploring its significance, the obstacles to its adoption, and strategies for building trust in AI technologies, providing guidance for leaders eager to integrate AI into their operations.

What is human-centred AI?

Human-centred AI is machine intelligence designed to enhance or transform how humans work, emphasising the symbiotic relationship between human intuition and machine precision. It seeks not to replace humans with machines, but to leverage the unique strengths of both – combining human creativity, emotional intelligence, and strategic thinking with the speed, accuracy, and data-processing capabilities of AI.

By placing humans at the forefront of AI adoption, this approach aims to enhance and redefine human interaction with technology. There are three key AI applications to the workforce that embody the principles of human-centred AI:

Automated intelligence

AI that handles routine, predictable tasks, freeing up human time for more complex and creative endeavours. This maximises the efficiency of machines in processing and executing tasks at scale while allowing humans to focus on areas where they add the most value. In our recent AI survey, 74 percent of leaders told us that their organisation would be more efficient if they better embraced AI.

Augmented intelligence

AI that enhances human productivity by providing insights, analyses, and recommendations that support faster and more informed decision-making. This application leverages the machine's ability to sift through vast amounts of data to augment human judgment and expertise, fostering a powerful partnership where each complements the other’s capabilities.

Additive intelligence

AI that unlocks new opportunities that were once beyond reach, expanding the limits of what humans can accomplish. Leveraging AI enables us to venture into uncharted territories across science, art, and innovation.

What are the key barriers to adopting a human-centred approach to AI?

Adopting human-centred AI poses unique challenges, primarily centred around the human aspect of the technology rather than its technical configuration. These challenges encompass the ethical deployment of AI, providing adequate support and skills for effective use of such tools, and addressing risks such as unintended outputs or biases.

Three key barriers to adopting human-centred AI include:

1. Uncertainty regarding the measurable impact of AI implementation

It is essential to clearly communicate the tangible advantages of AI, such as improvements in efficiency or enhanced decision-making capabilities, to encourage its adoption. For employees in organisations investing in AI, comprehending these advantages can significantly influence their support for the technology. The idea, for example, that AI could relieve them from the tedium of extensive data analysis and allow them more time for creative and imaginative value add activities, can be a powerful incentive to welcome AI initiatives. This transition not only aims to boost job satisfaction but also creates avenues for individuals to contribute to more meaningful and innovative work, fostering both organisational growth and personal development.

2. Underestimation of the necessary change management processes

The successful adoption of AI often demands considerable alterations to current processes, workflows, and interpersonal dynamics that are already in place. Recognizing and adequately planning for the magnitude of effort required to manage these transitions is vital.

3. Overlooking or inadequately addressing human biases and trust issues towards AI

Gaining trust in AI systems is a significant hurdle. Proactively addressing and making known biases in AI systems transparent, and emphasizing the importance of explainability, are essential steps in building trust.

Overcoming these challenges and embracing human-centred AI requires not just a bold mindset shift, careful experimentation, and a strong focus on involving and upskilling the workforce, but also the cultivation of trust as a foundational element.

Trust is essential for fostering a culture of innovation and for the effective navigation of changes brought about by human-centred AI. Building trust involves ensuring transparency, reliability, and ethical considerations in AI deployments, thereby reinforcing the confidence of both the workforce and stakeholders in the technology. By doing so, human-centred AI can truly maximise the respective capabilities and advantages of humans and machines, leading to unprecedented levels of collaboration and achievement. Establishing trust not only facilitates smoother transitions and acceptance of AI but also strengthens the synergy between human intuition and machine intelligence, paving the way for a future where both can thrive together in a shared ecosystem.

What are leading organisations doing to build trust in AI?

Leading organisations are adopting strategic approaches to build trust and foster responsible human-centred AI usage by aligning governance, management, and cultural practices. Here’s how this approach is simplified according to the "think big, start small, scale fast" principle.

Think big: Elevate AI management skills

Leaders must envision a comprehensive AI strategy where leadership accountability plays a crucial role. According to research from PA, 80 percent of leaders highlighted that the leadership team is responsible for AI implementation, providing necessary directives and guidance. This involves setting a clear vision for AI adoption within the organisation, underpinned by measurable objectives that ensure every AI initiative aligns with the overarching goals. Increasing AI management skills across the board is essential for this, involving tailored AI training for employees to cultivate a unified understanding of AI technologies.

Start small: Implement robust governance and foster a culture of experimentation

Initiating small-scale, strategic AI projects allows organisations to experiment responsibly with AI technologies. Governance plays a critical role in this phase, focusing on establishing processes and frameworks that are robust yet flexible enough to enable progress without compromising on ethical standards. Simultaneously, creating a culture of experimentation, where users are involved early in the development process, helps shape AI projects to align closely with user needs and priorities, mitigating negative externalities.

Scale fast: Accelerate adoption with strategic workforce planning and upskilling

Once foundational elements like governance structures and a culture of experimentation are in place, organisations can look to scale their AI initiatives quickly. This involves a focus on strategic workforce planning and a concerted effort to upskill employees, enabling the rapid scaling of successful Proof of Concept (PoC) projects. Understanding the Total Cost of Ownership (TCO) for AI initiatives is critical during this phase to assess the real cost-benefit impact over time effectively.

By following this structured approach, leaders can adopt human-centred AI in a way that is responsible, strategically aligned, and beneficial to both the organisation and its stakeholders, ensuring AI enhances rather than undermines human capabilities.