Loading blog…
Consultants have always worked in the grey areas where definitions shift, information sits across dozens of documents, and trust depends on evidence. It is a space built on human interpretation, structure, and judgement.
That is exactly why using AI right now in consulting work feels both promising and deeply frustrating. The technology is powerful, but not yet aligned with the way consultants think or the standards they need to maintain.
Most general purpose AI tools can write a good paragraph, but they cannot hold a consistent conversation about a complex client project. They can summarise a document, but they cannot explain how that summary connects to the rest of your materials. They answer quickly, but not always truthfully, and they rarely tell you where their answers come from.
At the same time, consulting firms everywhere are rapidly adopting AI, embedding it in research, drafting, and analysis. New roles are emerging to guide how consultants use it, and firms are rethinking how they structure teams and train staff. The question is no longer whether AI belongs in consulting, but whether the tools being used can actually meet the standards consulting requires.
So what do consultants need from AI?
Consultants need to be able to rely on the answers they receive. When the same question produces a different answer each time, it breaks the chain of reasoning that underpins consulting work. Clients expect a logical and defensible process, not an unpredictable system.
Most current AI tools, including widely used ones such as ChatGPT, do not guarantee consistent responses. The same prompt, entered twice, can produce different results. For creative writing this may not matter, but in consulting it does. A single shift in phrasing or emphasis can change the interpretation of an insight or the direction of an entire project.
AI must operate like a dependable colleague who remembers what was said yesterday. If a model cannot maintain that stability, every analysis becomes temporary and every insight fragile. Consistency is not just about accuracy, it is about credibility. Without it, consultants cannot reference outputs, build frameworks, or develop shared understanding within a team.
Consultants work in environments where every statement should be traceable. A claim is only as strong as its source. If an AI system provides an answer without showing where it came from, it creates uncertainty rather than clarity.
In recent months, several consulting reports produced with AI assistance have drawn criticism for including unverified or fabricated information. These cases underline how quickly reputational risk can surface when outputs are trusted without transparency and verification.
Transparency allows consultants to check reasoning, validate sources, and apply professional judgement. It means the AI should be able to point directly to a section in a report, a paragraph in a policy, or a dataset that informed its conclusion. In consulting, the difference between speculation and insight is the ability to show your work.
Most AI tools produce answers that sound right but are incomplete. They capture the surface of a problem, not its depth. For consultants, partial answers are dangerous because they distort the overall picture.
Completeness means being able to pull from all relevant materials, not just the most obvious or recent ones. It is the difference between quoting a single slide and understanding the entire deck. Consultants need AI that can hold multiple perspectives, recognise relationships across documents, and build responses that reflect the full context, not just the most convenient information.
Every organisation has its own structure of meaning, the definitions, categories, and relationships that shape how it reasons about its work. These definitions are part of its intellectual property and reflect how it understands itself.
Consultants work within these structures every day. Aligning on definitions is often what turns a collection of insights into a coherent strategy.
Most AI tools overlook this. They apply broad, generic meanings and blur distinctions that matter. True definition authority means AI must recognise and respect the internal ontology of an organisation, the web of concepts that gives its work structure and precision.
For consultants, this is essential. AI should fit into an organisation's way of thinking, not overwrite it.
Consulting projects are rarely one conversation. They unfold over months, sometimes years, through workshops, analysis, revisions, and reviews. AI that forgets or shifts direction over time cannot support that rhythm.
Stability means that the AI maintains awareness of earlier reasoning and continues to build on it logically. It should remember previous inputs, preserve decisions, and sustain a clear thread of continuity. Without that, consultants are forced to reframe the same questions repeatedly, losing the cumulative benefit that makes knowledge work efficient.
Consultants do not work with simple data. They work with detailed technical documents, layered stakeholder inputs, and information that changes across time. Most AI tools are built for short prompts and small contexts. They can summarise, but they cannot analyse at scale.
AI in consulting must handle complexity at the level of the work itself. That means processing large volumes of content, cross referencing structures, and managing hierarchies of information. It should be able to connect a policy clause with its related implementation report or trace a client objective back to the original strategic intent.
The next phase of AI, often described as agentic or autonomous systems, will raise the bar even further. As AI tools begin to plan, reason, and act across multiple steps, consultants will need to maintain even greater control and visibility over how those systems interpret their materials. Complexity will no longer be about the size of the data alone, but the autonomy of the technology.
Every consultant understands the importance of confidentiality. Client material is sensitive by nature and is often governed by strict data agreements. Many current AI tools are built on infrastructure where users have little visibility or control over how and where data is processed.
Consultants need sovereignty. They must know that data remains within trusted environments and that no external system can access it. This is not just a compliance issue but a matter of professional integrity. AI should empower consultants to apply their expertise safely, not force them to compromise client trust in exchange for convenience.
There is a quiet but critical expectation in consulting: when you do not know, you say so. AI must do the same. Fabricated answers erode trust faster than silence ever could.
Honesty about limits means that AI should be transparent about what it can and cannot infer from available data. It should make clear when evidence is missing or incomplete. Consultants can handle uncertainty, but they cannot work with invention. An AI that knows its boundaries is far more valuable than one that hides them.
These needs might sound obvious, yet most AI systems fail to meet them. They are designed for speed rather than depth, for style rather than substance. They produce answers that look confident but lack structure, provenance, or reliability.
Consultants require something different. They need a system that mirrors the discipline of their own methods: deliberate, grounded, and explainable. The gap between current AI and this level of trust is still wide, but closing it is what will define the next phase of intelligent consulting tools.
For AI to become a genuine consulting partner, it must think less like a writer and more like an analyst. It must connect evidence across documents, preserve definitions, and maintain reasoning over time. It must process complexity without losing precision and show its logic at every step.
When AI reaches that point, it will not replace consultants. It will amplify them. It will free time from searching, cross referencing, and verifying so that judgement and creativity can move to the forefront.
The consultants who gain the most from AI will not be those who prompt better, but those who demand more from the tools themselves.