Key insights from the ICA & LRN Masterclass: Managing Risk, Culture and Behaviour in the Age of AI.
Artificial intelligence has rapidly shifted from an emerging technology to an everyday business capability. Organisations across every sector are embedding AI into workflows, deploying copilots, experimenting with agentic AI and exploring how intelligent automation can drive greater efficiency.
But as AI adoption accelerates, governance is struggling to keep pace.
That was the central theme of a recent ICA & LRN Masterclass, where ethics, compliance and governance professionals came together to discuss one of the most pressing questions facing organisations today:
"How do we ensure AI enhances ethical decision-making rather than undermining it?"
The discussion made one thing abundantly clear: technical controls alone won't be enough. The organisations that succeed in the AI era won't simply have the most advanced technology, they'll be the ones that build cultures capable of governing it.
Adoption is moving faster than governance
One of the strongest themes throughout the session was the widening gap between AI adoption and organisational oversight.
Research shared during the masterclass shows that AI is already embedded across most organisations, yet governance frameworks remain immature. Employees are increasingly using AI tools independently—often without formal approval, training or clear organisational guidance.
This "shadow AI" isn't necessarily driven by malicious intent. More often, employees are simply trying to work more efficiently. If approved tools aren't readily available, many will turn to publicly available AI platforms to solve everyday business problems.
The challenge is that these seemingly harmless decisions can introduce significant risks around:
-
Confidential information
-
Intellectual property
-
Regulatory compliance
-
Data privacy
-
Model governance
The reality is straightforward: You cannot govern AI that you cannot see.
Visibility into AI usage has become just as important as deploying the technology itself.
AI governance is no longer just an IT responsibility
Historically, technology governance has largely been owned by IT, cybersecurity, legal and privacy teams. While those functions remain essential, the masterclass challenged attendees to broaden their thinking.
AI doesn't simply create technical risks. It creates behavioural risks.
The decisions employees make about when to trust AI, when to question it, when to disclose its use and when to override its recommendations are ultimately cultural decisions.
That means ethics and compliance teams have an increasingly important role to play alongside traditional governance functions.
Rather than asking: "Is this AI technically compliant?”
Organisations should also ask: "Will this AI encourage the behaviours and decision-making we want across our organisation?"
Culture is the governance layer that scales
One of the most powerful ideas explored during the session was that policies cannot anticipate every scenario AI will create.
Rules have limits. Culture doesn't.
Policies may define acceptable use, but they cannot dictate every judgement call employees make when using AI to draft reports, support investigations, analyse data or communicate with customers.
As AI becomes embedded into daily work, organisations need employees who instinctively ask:
• Is this accurate?
• Is this appropriate?
• Is this ethical?
• Should I verify this before acting?
Those behaviours cannot simply be mandated through policy. They have to become part of organisational culture.
In many ways, AI raises the importance of ethical culture rather than reducing it.
The greatest risk isn't hallucination, it's misplaced confidence
Much of today's AI conversation focuses on technical limitations such as hallucinations, bias and model accuracy. These are genuine concerns.
But the discussion highlighted another risk that may be even more significant: People stop challenging AI outputs because they look convincing.
One example discussed during the session involved a published report containing fabricated references that passed multiple layers of review before being identified externally.
The technology didn't fail in isolation: The review process failed and professional scepticism quietly disappeared.
As AI-generated content becomes increasingly polished, organisations will need to strengthen their culture of critical thinking. The future belongs to organisations where employees remain comfortable asking: "Is this actually correct?" Even when AI appears highly confident.
Treat AI agents like new colleagues
As organisations begin deploying increasingly autonomous AI agents, governance will need another shift in mindset. Rather than viewing AI solely as software, attendees explored the idea of treating AI more like a new member of the organisation.
Just as employees are:
-
onboarded,
-
trained,
-
coached,
-
assessed, and held accountable,
AI systems should be subjected to similar expectations.
That includes exposing AI to organisational values, codes of conduct and ethical standards; not simply testing whether it produces technically accurate answers. If AI is going to participate in organisational decision-making, it needs to reflect the culture organisations are trying to build.
Leadership will shape AI culture
The masterclass also challenged business leaders to think carefully about the messages they communicate. Employees are listening closely.
When leadership positions AI purely as a cost-saving exercise, it can generate uncertainty, resistance and fear.
When leaders position AI as a way to enhance capability, improve decision-making and support innovation, employees are more likely to engage responsibly.
That leadership message must also be backed by investment in:
• AI literacy
• practical training
• governance frameworks
• ethical guidance
• ongoing oversight
Building responsible AI isn't simply a technology project. It's an organisational change programme.
The future belongs to organisations that combine AI with judgement
Perhaps the strongest message from the session was that no organisation has completely solved responsible AI governance.
Everyone is learning.
The organisations most likely to succeed won't necessarily deploy AI first. They'll be the ones that preserve the qualities AI cannot replace:
-
human judgement
-
accountability
-
ethical reasoning
-
curiosity
-
professional scepticism
-
strong organisational culture
AI will undoubtedly transform the workplace. But technology alone will never determine whether organisations make better decisions. People - and the cultures they create - still will.
About LRN
LRN helps organizations build ethical cultures and more effective ethics and compliance programs through innovative education, behavioral science, culture measurement, and advanced analytics. Solutions such as LRN Reveal enable organizations to move beyond completion reporting to gain deeper insight into employee knowledge, behavioral trends, cultural indicators, and program effectiveness. By transforming compliance data into actionable intelligence, LRN helps organizations make smarter risk decisions and strengthen their culture of integrity.
