The Billion-Dollar Firms Are Hiring People, Not Buying AI
In February 2026, KPMG Australia quietly made a decision that sent shockwaves through the corporate world: they offshored 200 of their 260 executive assistant positions to the Philippines. Australian EAs earning $86,000 a year were being replaced by professionals at roughly $12,000 a year. The projected savings? $17 million annually.
Seven months earlier, Ernst & Young had done the same thing in the United States — laying off hundreds of domestic executive assistants and replacing them with offshore talent across the Caribbean and South America.
These are not scrappy startups. These are two of the four largest professional services firms on the planet. They have access to every AI tool on the market. Deloitte alone committed $3 billion to generative AI development through 2030. And yet, when it came time to solve the problem of high-cost administrative work, they hired people.
That should tell you something.
The AI Hype Trap
Every week, there is a new AI tool promising to replace your team. Automate your inbox. Automate your bookkeeping. Automate your customer support. The pressure to adopt AI is real — and for founders running businesses between $500K and $5M in revenue, it feels existential. Fall behind, and you are done.
But here is what the hype cycle does not tell you:
The tools change every week. Anthropic releases a new model. OpenAI counters. Google launches something new. A dozen startups pivot. The platform you bet on today might not exist in its current form six months from now.
For a founder doing $2M in revenue, this is not a $3 billion R&D problem you can absorb. A wrong bet on the wrong tool at the wrong time is not a rounding error — it is a quarter of wasted budget and distracted focus.
The question is not whether to adopt AI. It is how to adopt AI without burning money, time, and focus in the process.
The 4-Stage AI Adoption Framework
The most common mistake is trying to jump straight from your current state to full automation. It sounds efficient. It is actually the highest-risk path. Here is the framework that reduces risk while still getting you to AI — just on a smarter timeline.
Offshore Talent — Reduce Costs
Hire skilled offshore professionals to handle admin, research, data work, and operations. Get an immediate 60–80% cost reduction without disrupting your current workflow. This frees up your time and budget for what comes next.
Example: KPMG is saving ~$17M/year by offshoring 200 EA roles. A $2M founder can save proportionally — reclaiming 20+ hours per week.
AI Enablement — Augment Your Teams
Give both your offshore and onshore teams AI tools. Use AI to make people faster and more accurate — not to replace them. This is where you learn what AI can and cannot do in your specific business context.
Example: Your offshore researcher uses AI to summarize 50 industry reports in a day instead of 10. Same person, 5x output.
Gradual Automation — Automate What's Proven
After 6–12 months of AI-augmented work, you know what works. Automate only the workflows that have been validated by real usage — not by a vendor demo. Keep humans in the loop for judgment calls and quality control.
Example: Your team has used AI-assisted data entry for 8 months. Error rates are near zero. Now you automate the full pipeline.
Strategic AI Integration — Invest with Confidence
The AI landscape stabilizes. Clear winners emerge. You have real data on what works for your business. Now you make confident platform investments — full automation where it makes sense, human expertise where it does not.
Example: After two years of learning, you commit to a platform for your core workflow — knowing exactly what it will replace and what it will not.
Why This Order Matters
The instinct is to skip to Stage 4. Why pay for people when AI can do it? But this thinking ignores three realities:
First, you do not know what to automate yet. Most founders who jump to automation end up automating the wrong things — or automating processes that still need human judgment. Starting with people gives you the operational data to make informed automation decisions later.
Second, the AI market has not settled. Picking a platform in 2026 is like picking a search engine in 1997. Some of today's tools will dominate. Most will not. Stage 1 and 2 give you cost savings and productivity gains while the market sorts itself out.
Third, people are more flexible than software. An offshore professional can pivot to a new task in minutes. Retooling an AI automation pipeline takes weeks. When you are in growth mode, flexibility is worth more than theoretical efficiency.
The core reframe: This is not "outsource vs. automate." It is "outsource, THEN automate." Start with people. Layer AI on top. Let the market mature. Then invest with confidence.
The Evidence: What the World's Biggest Companies Are Doing Right Now
This is not a theoretical framework. The largest companies in the world are following this exact progression.
Offshored 200 of 260 EA roles to the Philippines. Savings: ~$17M/year. Phased rollout: April–June 2026.
Laid off hundreds of US executive assistants. Replaced with offshore talent in the Caribbean and South America. July 2025.
Meanwhile, Deloitte committed $3 billion to AI development while maintaining over 100,000 staff in India. All four major firms saw revenue declines in FY2024–25. Their response? Hire offshore talent for cost efficiency. Invest in AI for the long term. Run both in parallel.
The pattern is clear: people first, AI layered on top. Not because AI does not work — but because people give you the flexibility to figure out where AI works best.
What This Means for You
If the Big Four — with billion-dollar budgets and armies of technologists — are starting with offshore talent before going all-in on AI, there is a lesson here for every founder.
You do not need to solve the AI question today. You need to solve the cost and capacity question today. Get the right people handling your operational work at a fraction of the cost. Give them AI tools to multiply their output. Then, when the dust settles on the AI landscape, you will be in the strongest possible position to make strategic automation investments.
Stage 1 is available right now. The rest follows naturally.