By Jagdish Sharma – Co-Founder & CEO, Cedro
Most organisations are approaching responsible AI by asking an important question: What policies and regulations do we need?
But there is another question that deserves equal attention: Do our people have the skills and judgement to use AI responsibly?
Because that is where the use of responsible AI will come in.
As AI becomes embedded into everyday work from writing emails and analysing data to developing code, creating presentations and supporting business decisions the responsibility for how AI is used is no longer limited to technology, legal or compliance teams. It increasingly sits with every employee.
AI literacy is becoming workplace literacy
When computers entered the workplace, digital literacy became essential. As AI enters almost every function, AI literacy will follow a similar path.
But AI literacy cannot simply mean knowing how to write a good prompt.
Employees need to understand what information should never be shared with an AI tool, how to recognise that an AI-generated answer may be confidently wrong, how bias can influence an output, and when human judgement must override an AI recommendation.
The ability to question AI may ultimately become just as important as the ability to use it.
Consider some ordinary workplace situations. An employee uploads a customer document to summarise it. A recruiter uses AI to shortlist candidates. A developer accepts AI-generated code without fully reviewing it. A sales professional uses AI-generated market information in a customer proposal.
None of these actions may initially appear high-risk but each can create issues around confidentiality, accuracy, bias, intellectual property and even accountability.
A policy can tell employees what is permitted. It cannot make every decision for them.
From compliance training to everyday behaviour
This is why responsible AI needs to be treated as a skill rather than another annual compliance module.
Employees need opportunities to practise judgement through situations they are actually likely to encounter.
Can I put this information into an AI tool? Should I trust this output? How do I validate it? Could there be bias in this recommendation? Does this decision require human review?
These are practical skills, and like any skill, they improve through learning, practice and reinforcement.
The approach also cannot be identical for everyone. Responsible AI for an HR professional using AI in recruitment will look different from responsible AI for a developer, finance professional or customer-facing sales team.
Organisations thus need to move towards continuous and role-based AI learning, supported by real workplace scenarios rather than only policy documents and one-time awareness sessions.
Responsible does not mean restrictive
There is an equally important point: responsible AI should not become a reason for employees to fear or not use AI.
If governance is communicated only through restrictions don’t use this, don’t upload that, don’t experiment people may either avoid AI altogether or use it quietly without sufficient guidance. Neither outcome helps an organisation.
The objective should be responsible experimentation which includes creating clear guardrails while giving employees the confidence to explore, learn and innovate within them.
This will require organisations to bring technology, governance and learning together. Technology can provide controls. Policies can establish boundaries. But employees still need the judgement to navigate the situations that fall between those boundaries.
Over the next few years, organisations will spend significant time and money building their AI capabilities. The larger challenge may not be giving people access to AI. It will be building a workforce that knows how to use that access wisely.
Responsible AI cannot live only in a policy document.
It has to become part of how people think, decide and work every day.