AI and employee trust: Where internal communicators can create real value

AI is dominating workplace conversations right now, especially in internal communication and HR. Leaders are chasing speed and efficiency. Meanwhile, employees just want clarity, safety, and a bloody good reason to trust the change.

That is where the real tension lies. Tech adoption moves fast, but human trust moves slow. According to the IoIC index, only 36% of employees say leaders clearly explain how AI will be used, and just 35% believe it actually solves the right problems. At the same time, trust in leadership has dropped 9 points in a single year, sitting at a lukewarm 50%.

This gap matters because trust is the emotional contract between your leadership team and your workforce. Without it, introducing AI becomes a massive uphill battle. With it, organisations can move past the constant hype and start finding genuine value.

This blog post is a breakdown of everything covered in our AI, Trust and the Future of Work webinar by Jennifer Sproul, CEO at the IOIC, Dave Ferguson, Chief Technical and Product Officer at Oak Engage and Eva Armit, Customer Engagement Manager at Oak Engage. Let’s dive in.

Why trust sits at the centre of employee AI adoption

Trust drives followership. It dictates whether your people buy into big changes, understand the purpose behind them, and genuinely believe that leaders will listen, protect, and include them.

AI is arriving at a time when workplace trust already feels incredibly fragile. Public headlines constantly link AI with job losses, creating a heavy, nervous backdrop for any internal tech rollout.

When trust runs low, leaders are pushing water uphill. Employees start questioning motives, clarity, and their own place in the future of the company. AI does not create the trust problem, but it will expose it very quickly.

What the employee trust gap looks like in practice

The mismatch between leadership belief and employee reality shows up in the data:

• 49% of leaders believe AI is already part of everyday work, but only 42% of employees agree.
• 60% of leaders think employees are automating repetitive daily tasks, but only 36% of employees agree.
• 23% of CEOs believe teams are already delegating entire tasks to AI, but only 8% of employees agree.

It is a familiar pattern: leadership often believes adoption is miles ahead of where it actually is. Deployment is racing past trust.

Line managers are stuck right in the middle of this gap. Gallup’s latest research shows that the single strongest predictor of success in AI adoption is a direct manager actively championing it. Organisations see a far better chance of real transformation when managers lead the conversation.

This makes complete sense. People follow the attitude of their line manager long before they follow a generic broadcast message from the top.

How internal communicators can build trust around AI

Internal communication can turn AI from a scary threat into a shared, positive business change. But that has to start with a honest narrative.

Too many businesses roll out AI without a clear explanation of:

• Why they are using it
• What commercial benefits it supports
• What role it plays in the future of work
• What governance and compliance look like
• What it actually means for jobs, skills, and career paths

When you stay silent, employees fill the gaps with their own assumptions.

A stronger approach starts with honesty. Leaders need to explain what they know, what they do not know, and what the implications might be. It means involving employees in consultation and ideation early on, rather than announcing a finished tool and expecting immediate buy-in. Behavioural change does not happen overnight. It requires participation and inclusion.

A solid AI communication strategy should:

• Show the clear purpose of the change
• Explain the role employees play in it
• Give clear pathways for upskilling
• Use managers as the primary channel for trust
• Set firm governance from day one

If you skip this work, you end up trying to reverse-engineer trust later. That might give you a short-term win, but it causes long-term problems for company resilience and confidence.

Real AI value for communicators: less noise, more intelligence

AI creates the most value when it solves real, annoying problems instead of just churning out more content for its own sake.

The strongest use cases for communicators fall into three clear buckets:

1. Content creation

AI helps with drafting copy, making notes, and transcribing. Around 75% of communicators already use AI in this way, and it gives real, valuable time back to your day.

2. Self-serve support

Smart AI search helps employees find policies, benefits, and past announcements instantly, without waiting for a person to dig them out. This instantly reduces the daily admin pressure on comms and HR teams.

3. Analytics

AI turns messy data into useful insights. It helps teams understand platform usage, spot knowledge gaps, and shape future content strategy.

The real magic is not about doing more of the same. It is about using that saved time strategically.

Our profession is already facing a massive attention deficit crisis. Churning out more content does not solve problems—it just creates more digital noise. AI should help communicators cut through the noise, not add to it.

Think about the time wasted. Employees spend roughly 2 hours a day searching for information, and organisations use around 100 different software apps on average. AI can bring these disconnected systems together, reduce the friction, and free up teams to do work that matters.

But the big question remains: what are you doing with the time AI gives back?

Currently, only a third of communicators redirect that time into strategic planning, even though most admit AI helps them create content faster. That is a massive missed opportunity. AI works best when it frees you up to focus on high-value work.

Agentic AI: from responding to acting

Agentic AI shifts the technology from answering prompts to noticing what needs doing and taking action across systems.

In plain English:

• Generative AI answers a question
• Agentic AI notices a need, plans the steps, and takes action

Let’s look at a practical HR example. Instead of a human manually spotting a drop in team engagement, pulling a report, and emailing a manager, an agentic system can notice the dip, review the context, and suggest a check-in. A human stays in the loop for approval, but the system handles the heavy lifting of noticing and organising.

This opens up massive possibilities for internal comms and HR workflows. AI agents can handle monitoring and analysis, freeing people up to focus on judgment, interpretation, and real human connection.

Just keep your clarity intact. Giving AI agents human names can make internal conversations friendlier, but role clarity matters more. Keep a clear line between what the machine does and what the human owns.

Your company culture will impact whether people admit they use AI

At the events we’ve been attending this year, we’ve had a wonderful phone booth attached to our stand where comms professionals can leave their top secret, anonymous comms confessions. And one thing that became clear was that most communicators are already using AI far more than they let on. There’s a quiet embarrassment about it, like they’re cutting corners rather than just doing their job smarter.

When people fear negative consequences, they hide their tools. This creates shadow AI, siloed learning, and zero knowledge sharing. Psychological safety drops, and people stop speaking up. You end up with tiny pockets of hidden learning instead of open, cross-functional adoption.

The answer is not punishment. Threats just push the behavior underground. The stronger approach is to normalise the conversation.

Leaders can lead by example:

• Share exactly how they use AI themselves
• Be open about what they are still learning
• Explain where AI helps and where humans must stay in control
• Create casual spaces to share wins and failures
• Accept experimentation as a normal part of learning

Some forward-thinking organisations run AI show-and-tell sessions where people openly share what they tried, what saved time, and what completely failed. This type of openness makes AI feel like a normal part of organisational learning, rather than a confession.

It matters because 35% of employees are currently using their own money to pay for general AI tools. The demand is absolutely there, but it highlights a massive gap in corporate access, policy, and trust.

A practical test before any AI rollout

A strong AI initiative starts with a simple set of questions:

1. What specific, named problem does this solve?
2. Who says this problem actually exists?
3. What does success look like in 6 months?
4. Does the initiative still solve the problem over time?

Do not deploy AI before you understand the actual problem. In one recent case, 39% of business leaders made employees redundant as a direct result of deploying AI, and 55% of those leaders later admitted they got the decision wrong. The reason? A total lack of internal AI expertise.

Technology should never lead the process. Problem definition leads the process.

AI works best when you use it to analyse real issues, connect data across the business, improve workflows, and support better choices. That includes everything from customer feedback and employee sentiment to procurement, productivity, and process design.

AI adoption needs a human strategy

AI can absolutely help comms and HR teams save time, improve insights, and reduce daily friction. But the real value comes from how you use that extra time, the story you tell, and how clearly you involve your people in the journey.

The strongest AI strategies do not chase technology for its own sake. They solve real business and employee challenges, build genuine trust, and keep humans visibly in the loop.

When your communication stays clear, honest, and human, AI creates more than just speed. It creates space for better work.

Want to the watch the full webinar? Watch instantly on our webinar page, no email sign up needed.

‍