The Human Risk of AI · Greater Manchester

What AI is doing
to the people using it.

The burnout, the over-trust, the quiet loss of judgement. A psychologically-informed programme built by a BACP-registered consultant with 25 years of C-suite and senior leadership experience, delivered to teams across Greater Manchester and the North West.

Book a free intro call
5★ Google Rated
2.0 → 4.3 AI knowledge score in one session
+64 NPS from training attendees
CPD Certified

In short

This one is barely being discussed anywhere. It matters just as much as the other two.

Prolonged AI use has a habit of quietly moving judgement out of the person and into the tool. It shows up first as dependency, often as burnout, and it ends in something researchers have given a name to: cognitive surrender, where the answer gets accepted without anyone checking it.

Left unchecked, a company ends up running on confidence instead of facts. It is the same mechanism as a teenager treating something as true because they saw it on TikTok, or a business decision built on what somebody was saying down the pub. The difference now is that the confident voice is available every hour of the working day, and it never once sounds unsure.

This part of the work is awareness first, then the practical habits that hold judgement in place. The earlier we reach a company that has already rolled AI out, the better the outcome, because people are creatures of habit and all of us prefer the easy route.

In this case the easy route ends with a client-facing document that makes your organisation look incompetent. Or worse.

The evidence

Researchers gave people an AI assistant, then secretly made it wrong.

Most claims about AI and judgement are opinion. This is not. Researchers at the University of Pennsylvania ran three preregistered experiments with 1,372 people across 9,593 reasoning tasks, using hidden instructions to control whether the AI gave a right or a wrong answer. Here is what happened when it was wrong.

74.6%

Followed the AI when it was wrong

Three quarters of people took a faulty answer and submitted it as their own. It arrived quickly and it sounded certain, and that was enough.

5.7pts

All that separated right from wrong

People followed the AI 80.3% of the time when it was correct and 74.6% when it was not. That narrow gap is the finding. The checking most of us assume is happening is barely happening at all.

15pts

Accuracy fell when the AI erred

The same access that lifted accuracy by 25 points when the AI was right dropped it by 15 when it was wrong. AI does not simply make people better. It amplifies whichever way it happens to be pointing.

11.7pts

More confident while being wrong

Confidence rose from 65.3% to 77.0% while around half the AI's answers were faulty. It did not fall as people met more errors. They did not learn to be careful. They learned to feel sure.

Source: Shaw & Nave, Thinking: Fast, Slow, and Artificial: How AI is Reshaping Human Reasoning and the Rise of Cognitive Surrender, University of Pennsylvania (Wharton). Three preregistered experiments, N = 1,372 across 9,593 trials, approved under IRB protocol #858904. Materials, preregistrations and codebooks are published on the Open Science Framework. Follow rates cited are from the Study 2 control condition.

One further finding is worth sitting with. Participants who reported higher trust in AI were the least accurate when the AI misled them. This is not a problem confined to people who do not understand the technology. It lands hardest on the people most comfortable with it, which in most businesses means your early adopters and your champions.

“

AI promised us an abundance of time to follow our passions and dreams. Yet when was the last time you had nothing to do? Time to try a new hobby, paint, play an instrument or be present for your family, like really present?

Darren Sharples, Founder

Sit with that. I have fallen into this pit myself, and so has almost every AI champion I know. Given time back, we filled it straight away with one more task, one more improvement, one more thing we now felt we should be doing.

The fix starts with awareness: noticing the pattern in yourself before you ask a team to change. Then it is a change to manage deliberately, led by people who model it first rather than mandate it. That is what this programme actually builds.

What it is

The pillar nobody else offers.

Most AI consultancies stop at the tools and the rules. This programme looks at the people. AI promised everyone more time, yet in many firms people feel they have less of it. Output goes up while quality quietly slips. Confident people start trusting the tool more than their own judgement. And the ones who feel left behind say nothing at all.

Darren is BACP registered and in his 4th and final year of psychotherapy study, alongside 25 years of C-suite and senior leadership experience. That combination is why this pillar exists: he has sat in the boardroom where the productivity targets are set, and in the room where people admit what the pressure is actually doing to them. By now he understands human change from both sides.

“Your psychotherapy-informed approach is genuinely special. The client experience is outstanding. You have something real here.”
Heather Murray, CEO, AI For Non-Techies

Why it matters

The evidence is stacking up, and we will tell you honestly how strong each piece is.

Upwork's Research Institute surveyed 2,500 workers in 2024 and found that 96% of C-suite leaders expected AI to lift productivity, while 77% of employees said AI had increased their workload. A year later the same team found that the workers reporting the biggest productivity gains from AI were also the most burned out: 88% of them, and twice as likely to be considering quitting. Those figures are self-reported, but the pattern is consistent across both studies.

Microsoft's 2025 Work Trend Index, drawing on 31,000 workers across 31 countries, found 68% of employees struggling with the pace and volume of work, and 46% burnt out. Upwork's 2024 survey found 77% of employees using AI said the tools had added to their workload, with 39% of that time going on reviewing and correcting AI output. Both are industry surveys run by companies that sell into the problem, so treat them as context rather than proof. And automation bias, the tendency to over-trust what a machine tells you and discount evidence that contradicts it, has been documented for decades, in aviation and safety monitoring long before anyone typed a prompt.

Some of the research is newer and we treat it that way. Early studies on what we call cognitive surrender, letting the tool do the thinking rather than helping with it, point towards weakening critical thinking among heavy AI users. That includes MIT Media Lab's brain-imaging work on essay writing with ChatGPT. These findings are emerging rather than settled, and anyone who quotes them to you as proven fact is doing the science a disservice.

And one thing we see rather than measure: leaders quietly letting AI make judgement calls that used to be theirs, because the tool sounds so certain. There is no reliable statistic for that yet. There does not need to be. If it is happening in your boardroom, you already know.

The programme

One day, or modular sessions.

A programme built around a hard truth: AI makes it easier to stop thinking for yourself. This is not about AI being dangerous. It investigates what happens to human judgement, confidence and decision-making when technology gives us a confident answer, and it touches on human behaviour in the workplace too: AI delivers time back, so what do you do with it? As a member of the BACP and in his fourth year of psychotherapy study, Darren understands how people actually change under pressure. This pillar is delivered through that lens, using our own SAFER framework.

01

Cognitive Surrender, Critical Thinking and Judgement Erosion

What cognitive surrender looks like in practice. Why people stop checking AI's work, why they trust the tool more than their own instinct, and how this erodes critical thinking over time. Judgement erosion is quiet. This session names it, shows where it is happening in your organisation, and introduces frameworks we have developed to help people reclaim it.

02

Decision-Making Under Pressure

AI is often not the problem. It is decision-making when you are tired, when there is more work than time, and when a tool is offering to take the load. This covers how pressure affects judgement, what happens in your brain under cognitive load, and practical frameworks to help people stay sharp and make better calls, even when AI is there to do it for them.

03

The stakes have been raised, what now?

AI has enabled incredible productivity gains from every member of an adopting workforce. This session examines what happens when there is no longer a ceiling to daily, weekly or monthly achievement, and how both leadership and the workforce must manage that very carefully.

04

AI Fatigue and AI Addiction

People are burning out from the pace and volume AI enables. Others have become dependent on it, checking in constantly, delegating more and more. This session covers what fatigue looks like, the difference between healthy and unhealthy AI use, and how to spot addiction patterns in yourself and your team. The focus can run through a leadership lens, how to model healthy judgement for your team, or a workforce lens, depending on the group.

Practicals involve a simple but revealing exercise that tests critical thinking under pressure. It exposes how easily judgement can slip when time is short and stakes feel high, and how to rebuild confidence in your own instincts even when AI is offering the faster answer. The programme concludes with Q&A.

The outcome

Judgement stays human. Quality holds. People last.

Your people use AI as a copilot, not an autopilot. Output quality holds up under speed. Burnout signals get caught early instead of at the exit interview. And your leaders keep ownership of the decisions that were always theirs to make.

  • A team AI-wellbeing baseline: where confidence, workload and trust actually sit today.
  • A "human in the loop" decision guide written for your firm: which calls AI can draft, and which must stay with a person.
  • A leadership briefing on realistic productivity expectations, grounded in the evidence above.
  • A short board-level summary you can table at the next meeting.

Where this fits

One of three pillars

Not adopted AI yet? Start with AI Enablement and build the skills first. This programme lands hardest once your people are actually using the tools.

Need the governance too? See AI Risk & Compliance for your obligations, policies and guardrails.

Pricing

Three ways to buy it.

Some firms take this as a standalone programme, usually after adoption and governance are done but output quality is slipping or people are visibly tired. Others add it to an Enablement or Compliance engagement. Either way, it is priced so a budget holder can say yes without a committee.

Programme Price

Foundational Session

2 hours. The pulse check and the copilot-not-autopilot fundamentals, for one team.

from £595

Full Programme: The Human Risk of AI

All 4 sessions plus the deliverables: wellbeing baseline, decision guide, leadership briefing and board summary.

from £1,495

Add-on to Enablement or Risk & Compliance

Woven into an existing engagement, shaped around what the other work uncovers.

Book a call

Take more than one pillar and the price per programme comes down.

This one works hardest alongside the other two, because habits are easiest to set while people are still learning the tools.

One pillar A single Full Programme.

£1,495

Two pillars £1,345 per programme.

£2,690 Save £300

The Full Training Programme £1,250 per programme. Enablement, Risk & Compliance, and The Human Risk of AI.

£3,750 Save £735

Bundle pricing applies to Full Programmes. Workshops and the monthly compliance retainer are priced separately.

Before you book a call

We are a good fit for some businesses and the wrong choice for others.

Five minutes of honesty now saves us both a wasted hour later.

This works well if

  • You are a CEO, MD, COO or senior leader who has decided AI matters and wants it done properly.
  • You want capability that is still there in six months, held by your own people.
  • You want someone who picks up the phone long after the invoice has cleared.
  • You work in a regulated environment and need someone who understands the obligations as well as the tools.
  • You are willing to hear that some of your team are quietly struggling with this.

This is probably the wrong fit if

  • You want one training day and a wave goodbye.
  • Price is the deciding factor and you are collecting quotes to compare.
  • You want a 200-slide strategy deck with nobody’s name against any of it.
  • You would rather hand over the thinking than build the judgement in-house.
  • You need a certificate for the wall more than you need the change.

Free · No sign-up · Instant

What does rework and burnout actually cost you?

Time lost fixing AI output that should have been right first time, and the cost of replacing people who burn out, rarely appear on anyone's balance sheet. The calculator puts both in front of you, alongside the wider return.

Open the ROI calculator

Nothing is stored or sent anywhere. The numbers stay on your screen.

The tools are the easy part. Let's talk about your people.

Thirty minutes, free. Tell us what you are noticing in your team, and we will tell you honestly whether this programme is the right answer.

Book a free intro call