AI adoption without a strategy: why experimentation is not enough

24 Jul 2026

Posted at 19:18h in Blog, Communication by Murtuza Ratlamwala 0 Comments

AI is already finding its way into everyday work. Employees are using it to draft documents, analyse information, prepare presentations, summarise meetings and solve problems. In many organisations, this experimentation started before leaders had decided how AI should be used.

That creates an uncomfortable gap. AI activity is increasing, but the business may have no shared direction, clear priorities or consistent rules. People are using powerful tools, yet the organisation is not necessarily building measurable capability or value.

The challenge is no longer whether businesses should explore AI. It is whether they can move from scattered experimentation to a practical AI strategy.

AI activity is not the same as AI strategy

AI adoption can look impressive from a distance. Deloitte’s 2026 enterprise research found that workforce access to sanctioned AI tools increased substantially during 2025. McKinsey has also reported that generative AI is now regularly used in at least one business function by a clear majority of surveyed organisations.

However, widespread access does not mean AI is being used strategically.

The same McKinsey research found that fewer than one-third of organisations were following most of the practices associated with successfully adopting and scaling generative AI. Separate 2026 enterprise research found that 39% of organisations lacked a formal strategy for using AI to generate revenue, while many executives admitted their existing strategy provided little practical guidance.

This reflects what many organisations are experiencing. AI is being used, but use is fragmented. Different teams choose different tools, employees create their own methods and successful experiments remain isolated rather than becoming repeatable business practices.

What happens when employees move faster than the organisation

When there is no clear strategy, employees fill the gap themselves.

Some people avoid AI because they are uncertain about what is permitted. Others experiment with free public tools without understanding the privacy or confidentiality risks. More confident users develop useful ways of working, but those improvements may remain hidden within one role or department.

The result can be:

inconsistent quality and accuracy
sensitive information being handled inappropriately
multiple teams solving the same problem separately
investment in tools without clear business outcomes
uncertainty about who is accountable for AI-assisted work
a growing divide between confident users and everyone else

The organisation may appear active, but it is not necessarily moving in a shared direction.

Start with the business problem, not the latest tool

A practical AI strategy does not begin by asking, “Which platform should we purchase?”

It begins with questions such as:

Where are employees losing time on repetitive work?
Which processes create unnecessary delays or rework?
Where could better access to information improve decisions?
Which tasks require human judgement, and which could be supported by AI?
What result would show that an AI initiative is genuinely useful?

This shifts the conversation from novelty to value.

For example, the goal may not be “use AI in customer service”. It may be to reduce the time required to prepare a response while maintaining accuracy, tone and human oversight. That outcome can be tested and measured. A broad instruction to “use more AI” cannot.

The people strategy matters as much as the technology

AI adoption is often treated as a technology project. In practice, it is also a leadership, communication and capability challenge.

Employees need to understand why AI is being introduced, how it may affect their work and what responsible use looks like. Strong Communication & Interpersonal Skills Training helps organisations explain change clearly, encourage collaboration and build trust as AI becomes part of everyday work. Managers need enough confidence to discuss AI with their teams, assess potential applications and identify when human review is essential. Building these capabilities through Management & Leadership Training helps leaders guide AI adoption with confidence while supporting their teams through organisational change.

Without this support, uncertainty can turn into resistance, hidden use or overconfidence.

Deloitte’s 2026 research identified insufficient workforce skills as a major barrier to integrating AI into existing workflows. This is a useful reminder: organisations cannot purchase their way into AI maturity. People need opportunities to develop practical AI literacy, judgement and confidence.

A strong strategy therefore considers both the technology and the workforce that will use it.

Governance should make responsible use easier

Some organisations delay setting an AI direction because they are concerned about risk. Others respond with restrictive rules that employees do not fully understand.

Good governance should create confidence, not simply caution.

Employees need straightforward guidance on:

which AI tools are approved
what information must never be entered
when AI output needs to be checked
who remains accountable for the final work
how suspected errors or risks should be reported

Clear boundaries make responsible experimentation easier. They help employees understand where AI can add value without leaving them to make important privacy, accuracy and ethical decisions alone.

Move from isolated experiments to shared learning

Organisations do not need a perfect long-term plan before taking action. AI is changing too quickly for a strategy to remain fixed.

A more practical approach is to begin with focused, lower-risk use cases. Teams can test an application, measure the result and share what they learn before expanding it.

For each initiative, leaders should be able to explain:

the problem being addressed
the people involved
the expected benefit
the risks and safeguards
how success will be measured
what must be learned before scaling

This turns experimentation into organisational learning. Successful approaches can be repeated, while weak ideas can be stopped before they consume more time and money.

What a practical AI strategy should provide?

An effective strategy does not need to be a lengthy technical document. It needs to give people enough direction to make better decisions.

At a minimum, it should establish:

Purpose: Why the organisation is using AI and what outcomes matter.
Priorities: The business problems and use cases worth exploring first.
Guardrails: Clear expectations for privacy, accuracy, ethics and human oversight.
Capability: The knowledge and practical skills leaders and employees require.
Measurement: A way to assess productivity, quality, risk and business impact.

These elements create alignment without preventing innovation. And as AI is developing at a breakneck speed, make sure to regularly update your strategy.

How ICML can support the people side of AI adoption

Moving from AI experimentation to responsible adoption requires informed leaders, capable employees and a shared understanding of how AI should support the work.

ICML can tailor practical, in-house training to help organisations build this foundation. Workshops can help leaders and teams improve their AI literacy, identify useful workplace applications, understand responsible-use principles and develop greater confidence discussing AI opportunities and risks. Combined with Professional Effectiveness Training, employees can strengthen productivity, decision-making and practical workplace performance while using AI responsibly.

We also facilitate AI strategy sessions for executive teams, helping senior leaders explore the strategic implications of AI, identify high-value opportunities and develop a roadmap for responsible adoption and organisational impact.

The objective is not to chase every new tool. It is to help people make thoughtful decisions about where AI adds value, where human judgement remains essential and how responsible practices can be applied consistently.

AI adoption is already happening. The organisations that benefit will be those that turn individual experimentation into clear direction, shared capability and measurable workplace improvement.

Ask for a quote or request a proposal, and we will tailor an AI capability workshop to the priorities, risks and development needs of your organisation.

Explore moreon AI adoption to learn why clear strategy and governance are essential for successful workplace implementation.