The Eight Behavioural Risks of AI title

Artificial intelligence is developing at remarkable speed. Organisations are investing heavily in new tools, employees are finding new ways to use them, and regulators are increasingly focused on ensuring that AI is deployed responsibly. 

Much of the conversation has focused on technology, governance, and compliance. Less attention has been paid to the behavioural risks that emerge when people interact with AI in their day-to-day work. 

These risks are often difficult to identify because they do not arise from the technology itself. Instead, they arise from the way people think, feel, decide and behave when AI becomes part of the working environment. 

In his forthcoming book, Artificial Intelligence and Behavioural Risk: The Danger and the Saving Power, Philip Smith (from Nicholson McBride Fisher) identifies eight key areas where AI can create vulnerabilities for individuals, teams and organisations. 

1. Cognition 

Can we distinguish reliable evidence from plausible appearance? 

AI can produce convincing answers in seconds, but confidence and accuracy are not the same thing. The challenge is maintaining critical thinking, recognising uncertainty and checking important information before acting on it. 

2. Emotion 

Can we respond appropriately to simulated emotion? 

As AI becomes increasingly capable of mirroring empathy, encouragement and concern, people may find themselves reacting emotionally to systems that do not genuinely understand them. This raises important questions about influence, dependency and judgement. 

3. Identity 

Can we maintain an authentic sense of who we are? 

When AI contributes to our writing, communication and decision-making, the boundaries between human and machine-generated work can become blurred. Organisations must consider what authenticity means in an AI-supported world. 

4. Skill 

Can we retain essential human capabilities? 

One of the most frequently discussed risks is overreliance. When AI performs tasks that previously required expertise, there is a danger that knowledge, judgement and capability begin to erode over time. 

5. Influence 

Can we challenge AI-mediated decisions? 

AI increasingly shapes what we see, prioritise and act upon. The question is whether individuals feel able to question recommendations, challenge assumptions and exercise independent judgement. 

6. Privacy 

Can we maintain appropriate boundaries around information? 

The convenience of AI can encourage people to share information without fully considering how it may be stored, processed or used. This creates both organisational and personal risks. 

7. Ethics 

Can accountability remain clear? 

When decisions involve AI, responsibility can become blurred. Organisations need to ensure that actions remain open to scrutiny, challenge, correction and accountability. 

8. Trust 

What makes AI worthy of trust? 

People often trust technologies because they are familiar, authoritative or convenient. Genuine trust, however, should be based on demonstrable competence, transparency, integrity, and an understanding of context. 

Looking Beyond Technology 

These behavioural risks are not arguments against AI. On the contrary, AI has enormous potential to improve performance, productivity and innovation. 

The challenge is ensuring that organisations understand the human vulnerabilities that accompany these benefits. By recognising the early warning signs and developing effective habits and safeguards, firms can maximise the value of AI while reducing unintended consequences. 

As AI becomes more deeply embedded in organisational life, understanding behavioural risk may become just as important as understanding the technology itself.