The Consultant Who Built the AI That Displaced Him

Two hours after Donald King won first place in PwC's firm-wide OpenAI hackathon, he was laid off.

He was 26, three years into what he called his dream job. For the previous eight months, he had worked 60 to 80 hours a week at PwC's AI factory, building fleets of AI agents that automated manual tasks for Fortune 500 clients. He organized internal knowledge-sharing events that drew 250 colleagues. He won the hackathon. Then, in October 2024, the firm let him go. "I thought I was safe," he said later. "I just got a little blindsided."

The part that stayed with him wasn't the timing, as brutal as it was. It was the conclusion he reached afterward: the AI agents he built may have contributed to staffing cuts at the clients he served. "It's 100% connected," he said. He had been building the technology of his own industry's disruption while working inside one of its most prestigious firms — and proximity to that technology hadn't protected him.

If you work in consulting right now, King's story is worth sitting with for a moment. Not because it's your story. But because it sharpens the actual question: not whether AI is changing consulting, but which parts of your specific work are already being handed off to a model, which parts are becoming more valuable than ever, and what you can do about it before circumstances force the issue.

But King's story is more complicated than a cautionary tale — and that complexity is where the useful information lives.

The Disruption Is Real, and It's Structural

The scale of what's happening inside consulting firms isn't abstract anymore. Entry-level consulting job postings have collapsed 35% since the start of 2023. McKinsey shed roughly 5,000 employees from its peak headcount of 45,000, and in late 2025 signaled further cuts of up to 10% in non-client-facing roles. PwC laid off 1,500 employees. KPMG cut 33% of its graduate cohort.

The economic model underneath these headcount decisions is also shifting. KPMG's US Vice Chair of Advisory, Rob Fisher, said billable hours are already gone in "a really large portion" of the business — not a prediction, but a description of current operations. McKinsey now generates roughly a quarter of its fees from outcome-based pricing. The traditional consulting pyramid, in which many juniors do research and slides while a few seniors do strategy, is compressing.

That picture sounds alarming. The context that matters: the International Council of Management Consulting Institutes puts the automation likelihood for consulting at 27% — of tasks, not jobs. Less than 5% of consulting roles can be fully automated. The profession isn't disappearing; its internal economics are being renegotiated, and the renegotiation is happening fastest at the junior end.

This pattern extends well beyond management consulting. Goldman Sachs data shows 72% of AI-driven displacement is hitting white-collar roles — the inverse of every previous automation wave. If you're in marketing strategy, HR advisory, financial analysis, or content development, the same structural forces are reshaping your profession's junior layer simultaneously. The question is the same regardless of your job title: which parts of what you do on Monday morning are already at risk?

Which Parts of Your Work Are Already Gone

This is where the useful answer lives, and it's more specific than most AI coverage suggests.

A Harvard Business School study of 758 BCG consultants found that AI-enabled consultants completed tasks 25.1% faster and produced output rated more than 40% higher in quality. The tools are genuinely impressive — for the tasks they can handle. Market research and competitor scanning. First-draft slide and report production. Standard financial modeling. Regulatory and data synthesis. These are the tasks migrating to AI inside consulting firms right now, and a Stanford study confirmed the downstream effect: early-career workers aged 22 to 25 in AI-exposed roles saw a 16% relative employment decline between 2022 and 2025.

The tasks gaining value look different. Client trust and relationship management. Organizational navigation and change implementation. Ethical accountability and AI governance. Strategic framing under genuine ambiguity. These are the tasks where AI, per the research, cannot bear responsibility, read room dynamics, or substitute for the judgment that comes from having been inside difficult organizational situations before.

Left-brain thinking — analysis, synthesis, structured outputs — is going to be commoditised by AI. Right-brain thinking — creativity, connection-making, navigating ambiguity — that's where differentiation lives.
— Soren Kaplan, Innovation Consultant

The most important piece of evidence here comes from a separate HBS field experiment. Researchers tested an AI assistant among entrepreneurs in Kenya, and found that AI boosted top-performing entrepreneurs by 10 to 15% while actually decreasing outcomes for weaker performers by roughly 8%. The finding: AI amplifies existing judgment rather than replacing it. Bring strong judgment to AI and it compounds. Bring weak judgment and it accelerates your mistakes.

Run the audit on your own last two weeks. What percentage of your hours went to the first list versus the second? The answer is a more honest career assessment than any job-security survey.

This task spectrum applies across professional knowledge work, not only consulting. Marketing strategists face the same divide between AI-generatable content briefs and AI-resistant client relationships. HR advisors face it between automated policy drafting and the genuinely human work of managing conflict or designing culture. The specific tasks differ; the underlying pattern doesn't.

What the People Who Navigated This Actually Did

Understanding where the risk is concentrated tells you something important. The more actionable question is what the people who have navigated this transition successfully did differently — and whether that path requires being exceptional.

Julius Bruch spent seven and a half years at McKinsey — as a physician-scientist with an MD and a PhD in dementia research — before leaving to found Isaac Health, an AI healthtech startup focused on brain health and dementia care. The company raised $16.3 million. The detail that earns trust isn't the funding number. It's the doubt that preceded it.

"Initially, it felt a bit like: What am I doing? Am I just playing around?" he recalled. He gave himself a three-month deadline, and when enough traction arrived inside that window, he kept going. What he had to unlearn from McKinsey surprised him: "The main thing to unlearn from consulting was the pursuit of perfection." In a startup — and in any AI-augmented work context — the time cost of exhaustive analysis before action is no longer offset by the quality it produces. AI handles the exhaustive part. What it can't do is decide when good enough is enough and move.

The main thing to unlearn from consulting was the pursuit of perfection.
— Julius Bruch, Cofounder and CEO, Isaac Health

Four other former MBB consultants who made similar transitions told Business Insider the same thing: they all had to abandon the habits that had got them promoted. Not because those habits were wrong, but because the value equation had changed.

This habit disruption isn't unique to McKinsey. Any professional trained in institutions that reward thoroughness over speed — law firms, corporate strategy teams, academic environments — faces the same recalibration when AI changes the cost of getting to a first draft. The consulting instinct to stress-test every recommendation before presenting it was a strength when research took days. When AI can produce a defensible first-pass analysis in minutes, that instinct becomes a bottleneck.

Bruch's story doesn't mean leaving your firm is the right move. It means the transition from "AI is threatening my work" to "AI is working for me" is primarily a habits question, not a tools question. The two or three behaviors most optimized for the old model are worth identifying now, before external pressure makes the choice for you.

What to Do Before the Week Is Out

King's TikTok post about his layoff got 2.1 million views. Most of the comments expressed shock. What gets less attention is what happened in the 60 days that followed. He turned down every corporate job offer he received, launched his own AI agency — AMDK — and within a year was serving clients ranging from small companies to billion-dollar enterprises. "I'm grateful for it happening," he said. "It was the worst thing that ever happened to me, but then it turned into the best thing."

That arc isn't a prescription. Not everyone who gets laid off launches an agency in two months. But it is a data point about what the window after disruption — or before it — can look like when the person going through it has domain knowledge and a specific plan rather than a vague intention to "upskill."

The AI consulting market is growing at 26.5% annually. Demand for consulting isn't disappearing — it's moving, from research and slides toward implementation, governance, and the judgment-intensive work that AI cannot do accountably. The consultants capturing that demand aren't necessarily the most technically sophisticated. They're the ones who audited their own task lists honestly and started shifting their time before circumstances forced it.

Three actions that require no external platform or course enrollment:

First, run a task audit. List every deliverable you produced in the last two weeks. Mark each one: "AI could first-draft this" or "AI couldn't responsibly own this." The ratio is your personal exposure score — and it's more useful than any industry survey.

Second, make one deliberate handoff. Identify a deliverable from the first list and actually use an AI tool to generate the first draft next time. Note where your judgment improved it. That gap — between what AI produced and what you knew needed to change — is what you're actually selling.

Third, lead with judgment in your next client conversation. Rather than opening with a prepared output, lead with a question that requires contextual reading: organizational politics, unstated constraints, emotional dynamics. Practice the skill that compounds with AI rather than competes with it.

The consultants who thrive through this won't be the ones who avoided the disruption. They'll be the ones who stopped treating it as a general threat and got specific about what it meant for Monday morning.

Recommended Tools & Resources