My AI Burnout theory begins with a warning: we can automate work faster than we can redesign the human responsibility around it.
AI can accelerate production, multiply options and keep work moving around the clock. People remain accountable for understanding the outputs, making decisions and dealing with the consequences.
When those responsibilities expand without sufficient authority, support or recovery, pressure builds.
I believe AI Burnout will become one of the defining leadership risks of the Human x AI era. It belongs on the board agenda.
It also changes what we should demand from executive coaching. Leaders navigating this pressure need practitioners whose competence has been assessed, whose conduct is accountable and whose methods can withstand scrutiny.
That is why I see certification, governance and effective regulation as essential to the future of coaching.
My AI Burnout theory proposes that sustained exhaustion can emerge when AI increases the demands placed on people faster than organisations increase their capacity to understand, influence and manage those demands.
At its centre is a dangerous imbalance:
Responsibility expands while meaningful control contracts.
A leader can generate more strategic options but has less time to evaluate them. A manager oversees more activity but has limited visibility into how the work was produced. An employee becomes faster at completing tasks but the requirement to self asses and increase in output requirement immediately happens.
In each case, greater technical capability can coexist with less human control.
The theory brings together four pressures: the need to keep learning, the obligation to verify machine outputs, uncertainty about professional value, and the expectation of continuous availability. My proposition is that these pressures become particularly damaging when people lack the authority to challenge the demands being placed on them.
I am developing AI Burnout as an organisational theory, open to research, testing and refinement. Its central challenge to leaders is already practical: who carries the pressure created by your AI strategy, and what power do they have to manage it?
Microsoft’s 2026 Work Trend Index found that 65% of the AI users surveyed feared falling behind if they failed to adapt quickly, while only 26% said their leadership was clearly and consistently aligned on AI. These findings do not establish my theory, but they illustrate the uncertainty in which people are being asked to perform. Microsoft, 2026 Work Trend Index
For me, the response requires three connected layers of accountability.
1. Governance must assign responsibility for the human consequences of AI.
Boards and executive teams should know how AI changes workload, decision authority, professional development and expectations of availability. Someone must be accountable for examining those effects and acting on them.
A business case that promises faster output should also explain who will review it, how much capacity that requires and what happens when the system exceeds that capacity. Productivity measures should sit alongside evidence about decision quality, rework, working hours and people’s ability to raise concerns.
My theory makes a specific prediction worth testing: comparable AI deployments will create different levels of strain depending on how well responsibility, authority and support are aligned.
2. Management controls must give people the power to intervene.
A person held accountable for AI-assisted work needs time to scrutinise it, access to relevant expertise and authority to stop or escalate it.
Organisations should establish realistic review capacity, clear escalation routes, protected learning time and explicit expectations about availability. Managers should decide how time saved will be allocated before it silently becomes another increase in workload.
In the context of AI Burnout, control means being able to influence the demands you are responsible for meeting. Asking someone to approve an output they cannot adequately assess creates accountability without the conditions needed to exercise it.
3. Professional regulation must make those supporting leaders answerable for their practice.
Coaches can influence how leaders interpret pressure and respond to it. That gives the coaching profession a responsibility to establish credible entry standards, ethical boundaries and effective oversight.
Professional self-regulation already provides mechanisms for this. ICF’s model includes competency standards, credentials, ethical review and continuing education. Those mechanisms apply within its professional remit; they are distinct from universal statutory licensing of everyone offering coaching.
I want stronger, more consistent protection across the industry: verifiable credentials, accessible complaints processes, independent investigation and meaningful consequences for misconduct. Where voluntary oversight leaves significant gaps, the profession should engage seriously with the case for statutory intervention.
These three layers meet in the coaching relationship.
Consider a leader who says: “My team is producing more, but everyone seems overwhelmed.”
A competent coach should be equipped to examine the expectations, incentives and decisions behind that situation. Has AI increased the volume of work requiring approval? Have targets risen without a review of capacity? Can employees challenge the pace? Is the leader rewarding constant availability?
If the conversation stops at helping individuals become more resilient, the conditions creating the pressure may remain untouched.
Coaching must be capable of challenging the system that produces the strain.
This is why certification is essential to my argument.
A coach working with these issues needs assessed skills in listening, challenge, ethical judgement, contracting and professional boundaries. They must understand the limits of their competence and recognise when a client needs specialist support. They must also examine their own assumptions about productivity, performance and success.
My position is clear: recognised, assessed coaching qualifications or professional credentials should be a minimum requirement for professional executive coaching appointments.
That requirement should be supported by relevant experience, ongoing supervision and continuing development. Organisations need evidence that the practitioner can apply their learning responsibly to the work being commissioned.
They must also scrutinise what a credential actually means. Programme accreditation, an assessed qualification and an individual professional credential provide different forms of assurance. ILM offers regulated coaching and mentoring qualifications; ICF distinguishes education accreditation from individual credentialing. Buyers should verify the specific award and what it demonstrates.
Certification cannot guarantee that every intervention will be effective. Its importance lies in establishing an assessed foundation that can be reinforced through oversight, supervision and accountability.
Clients also need somewhere to turn when standards fail. ICF’s Ethical Conduct Review process provides a route for complaints against its members and credential-holders, with consequences that can include loss of membership or credentials under its procedures.
Professional influence should carry professional consequences.
The same principle applies when coaches use AI themselves.
A practitioner exploring AI Burnout with a client should be able to explain their own use of technology: what information is processed, what the client has agreed to, how outputs are checked and who remains responsible. The ICF’s AI coaching framework addresses areas including ethics, privacy, bias and assurance. These questions should be part of commissioning and reviewing coaching services.
Employers must provide the surrounding governance. Every coaching programme should have clear ownership, verified practitioner standards, agreed objectives and a route for raising concerns. Reporting arrangements should be established in advance so that organisational accountability protects the confidentiality needed for honest conversation.
This is what Human x AI means to me: greater technological capability matched by stronger human agency.
People should understand the work they are responsible for. They should have the authority to question it, the support to develop their judgement and the capacity to recover. Leaders should be accountable for creating those conditions.
My AI Burnout theory connects these responsibilities. It asks us to examine where pressure is created, how it is transferred and whether the people carrying it retain meaningful control.
For 2027, I want that question embedded in leadership development, coaching practice and board oversight.
At Actuate, this is the standard I want us to advance: qualified coaches, accountable leaders and organisations prepared to govern the human consequences of their ambition.
If we give technology more power, we must give people stronger protection, clearer authority and better-qualified support.
By Matthew Michael Roberts, CEO
