Answer capsule
An executive should leave coaching with greater ownership of the decision, not dependence on a coach, script, or AI platform that quietly chooses the agenda and answer.
What the source establishes
- ICF says its current Core Competencies framework was introduced in 2025 after a review process that included a job analysis of more than 3,000 coaches across disciplines and backgrounds.
- The framework is organized into four domains: Foundation, Co-Creating the Relationship, Communicating Effectively, and Cultivating Learning and Growth.
- ICF describes the competencies as a framework for professional coaching practice and as a basis for credentialing and accredited coach education.
- The ICF page connects coaching with client ownership of development, client-set goals and actions, accountability, and autonomy.
Define the decision the executive must continue to own
Before comparing coaches, name the leadership decision, behavior, or transition the executive wants to work on and the authority that must remain with the executive. Clarify what a useful engagement would help the client notice, examine, choose, practice, and review. ICF's current competency model emphasizes a client-centered relationship, client ownership, goals, action, accountability, and autonomy. Those ideas create a practical buying test: the coach should support the executive's thinking and learning without taking over the decision or making continued progress depend on the coach's preferred answer.
Put that boundary into the engagement agreement. Identify who sets goals, how the agenda can change, what the coach will challenge, what information the coach may introduce, and how actions are selected. If an employer sponsors the work, state which objectives belong to the sponsor and which remain the client's. If an AI tool proposes questions, exercises, or next steps, disclose that role and preserve a way for the executive to reject or redirect it. Autonomy is not the absence of challenge; it is the client's informed authorship of the judgment and action that follow.
The accountable team should translate this point into a named workflow, affected population, source data, human owner, approval right, exception path, retained evidence, and review date. That translation is what separates an interesting AI development from a decision that can be governed and evaluated.
Observe whether the method expands or narrows choice
A credential, biography, chemistry call, or branded framework does not show how the coach behaves when the executive is uncertain or disagrees. Ask the candidate to explain how sessions are contracted, how client language is reflected, how assumptions are tested, how progress is reviewed, and what happens when the coach's hypothesis is wrong. In a sample conversation, notice whether questions open inquiry or steer toward a preselected conclusion, whether silence and dissent are allowed, and whether the executive can define the meaning of success rather than accept a generic score.
Evaluate technology in the same way. A platform can support reflection and follow-through, but it can also rank topics, optimize for engagement, recommend a path, or turn a sponsor's priorities into default prompts. Ask who designed those choices, what evidence supports them, what participant data changes them, and whether the named coach can override the system. Preserve examples of actual session and product behavior. ICF's public description expresses professional expectations and promotional claims about coaching; it does not independently verify that a particular practitioner or platform follows them.
The accountable team should translate this point into a named workflow, affected population, source data, human owner, approval right, exception path, retained evidence, and review date. That translation is what separates an interesting AI development from a decision that can be governed and evaluated.
Separate coaching from advice, training, and implementation
An executive may legitimately need consulting, mentoring, therapy, legal advice, technical instruction, facilitation, or implementation support in addition to coaching. The fit problem arises when those services are blended without a clear role change. Ask the provider to name which mode is being used, whose expertise drives the answer, what duty or credential applies, how conflicts are handled, and how the client will recognize the transition. For AI decisions, a coach can help an executive surface tradeoffs and accountability without pretending to be a cybersecurity, employment, procurement, privacy, or engineering authority.
If the provider also sells an AI product, implementation service, assessment, or training program, record the commercial relationship and the alternatives the executive can choose. A recommendation to buy the provider's adjacent service should not be presented as a discovery generated solely by the client's reflection. Re-contract before shifting roles and document what information transfers between them. Clear role boundaries protect autonomy because the executive can evaluate advice as advice, coaching as coaching, and product claims as product claims rather than treating one professional label as evidence for all three.
The accountable team should translate this point into a named workflow, affected population, source data, human owner, approval right, exception path, retained evidence, and review date. That translation is what separates an interesting AI development from a decision that can be governed and evaluated.
Review autonomy without claiming the outcome
At agreed intervals, ask the executive for concrete evidence: Can the client state the decision more clearly, identify assumptions, generate alternatives, choose an action, explain the basis, and continue the work without the coach present? Review whether goals and actions remain client-owned, whether challenge is useful, and whether the relationship increases or suppresses candor. Also inspect missed sessions, dependency signals, repeated formulaic outputs, pressure to renew, and sponsor reporting that may pull the agenda away from the client's agreement. These observations can support an engagement decision without turning a subjective experience into a universal performance claim.
The 2025 competencies were informed by a broad coach job analysis and are used in ICF credentialing and education, but they are not a fit score, product certification, clinical standard, or guarantee of leadership change. A buyer should verify the individual coach's current credential and conduct, the exact service and technology, the sponsor arrangement, and any adjacent professional obligations. Client autonomy is therefore a diligence lens: it helps an executive ask whether the engagement preserves ownership, while leaving outcome claims to evidence with a defined population, method, comparison, and limitation.
The accountable team should translate this point into a named workflow, affected population, source data, human owner, approval right, exception path, retained evidence, and review date. That translation is what separates an interesting AI development from a decision that can be governed and evaluated.
Decision test
Ask whether the source changes the decision itself, the evidence required, the implementation sequence, or only the language used to describe an existing capability. Record which claims are directly supported, which are provider statements, which require an independent test, and which remain unknown. A source-linked review should make uncertainty easier to see, not bury it inside a blended score.
Questions to take into review
The publication supports research and executive decision preparation. It does not provide legal, financial, accounting, employment, clinical, cybersecurity, investment, procurement, or implementation advice.