Microsoft has a talent picture the market should be watching closely. They're the largest tech employer in this comp set at 210K, but they're the only one shrinking — down 6% year-over-year.

It's worth noting that headcount reduction doesn't automatically signal weakness — Microsoft may be becoming more efficient through AI tooling and automation. But the 18% attrition rate suggests this is at least partly involuntary churn, not planned optimization.

Microsoft's 18% attrition rate (Big Tech avg: ~11-12%, est.) is the highest of the Big 4 by a wide margin — up from ~14% (est.) in Q2 2025, meaning the bleed is accelerating. Google is at 11%, Apple at 10%. Even Amazon — famous for churn — is only at 16%. Microsoft is bleeding talent faster than anyone.

And the hiring numbers don't offset the losses. Despite 24,043 hires in the last 12 months, the net headcount is still falling. They only have 2,551 open roles — the fewest of the group, down from ~4,200 (est.) six months ago — which suggests this isn't aggressive pruning with plans to rehire. Compare that to Amazon's 13,718 open jobs or Google's 5,309. Microsoft appears to be in contraction mode.

This is where it gets really uncomfortable for the bulls. The talent data does NOT support the AI narrative.

Look at Microsoft's fastest-growing skills: Elasticsearch (+4%), AWS (+4%), Data Structures (+2%), Docker (+2%), NoSQL (+1%). These are generic infrastructure skills, not AI skills. There is not a single AI or ML skill in Microsoft's top growing skills list. And even their peak growth rate of 4% has been flat since Q3 2025 (est.) — no improvement despite the Copilot push.

Now compare that to Apple — a company nobody talks about as an AI leader. Apple's fastest-growing skills are Deep Learning (+12%), Machine Learning (+11%), Applied ML (+11%), and NLP (+10%). Apple is quietly building the strongest AI talent bench in Big Tech.

Google is growing React.js (+17%), NoSQL (+14%), Elasticsearch (+13%), MongoDB (+13%), Docker (+13%) — high-velocity skills growth across the board. Microsoft's peak skill growth is 4%. Four percent.

The AI narrative around Microsoft may be driven by the OpenAI partnership and marketing, but the workforce data tells a different story. They aren't building the AI muscle internally at anywhere near the rate you'd expect.

Microsoft is decisively losing the talent war to Google. This is the single most bearish data point in the entire analysis.

Microsoft lost 711 people to Google in the last 12 months while hiring only 268 back — the net loss of 443 has widened from ~310 (est.) in the prior 12-month period. That's a net loss of 443 engineers to their most direct competitor. Google's #1 source of incoming talent is Microsoft — they are literally Microsoft's biggest brain drain.

The Amazon picture is more favorable. Microsoft hired 1,187 from Amazon and AWS combined (660 from AWS, 527 from Amazon), while losing 401 back. That's a net +786 — but this is largely a Seattle labor market effect, not a strategic win.

Microsoft is also losing heavily to Meta (368 out, no significant inflow), to Oracle (345 out), and even to HCLTech (279 out). The outflow to HCLTech (279) is notable — it suggests at least some of the departures are to services firms rather than competitors.

Meanwhile, Apple is only losing 151 people to OpenAI — its biggest outflow. Microsoft doesn't even show up as a significant talent threat to Apple.

Microsoft scores 41/100 — lower half of our 24-company dataset. For context, here's how the Big Tech peers compare:

Microsoft's weakness is concentrated in Retention (5/20 — 18% attrition, worst in the group) and Hiring Intent (6/20 — only 2,551 open roles). Pedigree is moderate at 10/20 — the employer quality component is solid (AWS, Amazon, Google, and Salesforce are all top inflow sources), but the school component is weaker than peers. Google scores 16/20 on pedigree, pulling from Microsoft, Amazon, AWS, and Meta — and with 5.7% of its workforce from elite schools. Apple scores 14/20 with Amazon, Google, Microsoft, and Meta as inflow sources. Microsoft's pedigree story is that it attracts from the right companies, but its elite school concentration doesn't match Google or NVIDIA. The gap to NVIDIA (79) and Google (70) is substantial.

Talent signal: Negative across most dimensions. The talent data contradicts the market narrative on several fronts.

Here's what the workforce signals show:

Bear signals (strong):
- Headcount shrinking -6% YoY (only shrinking hyperscaler; Big Tech avg: ~+5%, est.)
- 18% attrition rate (Big Tech avg: ~11-12%, est.) — highest in the comp set by a wide margin, nearly double Apple's 10%, and rising from ~14% (est.) in Q2 2025
- Losing the talent war to Google badly: net -443 people, Microsoft is Google's #1 hiring source
- Losing significant talent to Meta (-368) and even HCLTech (-279)
- Only 2,551 open roles — lowest of the four, signals no intention to backfill
- Zero AI/ML skills appearing in fastest-growing skills — peak skill growth is just 4%
- The AI narrative is not showing up in the workforce data at all

Lone bull signals (weak):
- Net talent winner against Amazon (+786), but this is largely the Seattle labor market gravitational pull, not a strategic edge
- Employer pedigree is solid: AWS, Amazon, Google, and Salesforce all feed Microsoft's talent pipeline — four elite employers in the top inflow sources. But strong inflows don't matter if the back door is open wider

The market is pricing Microsoft as an AI leader on the strength of the OpenAI partnership and Copilot branding. But when you look at what the engineers are actually doing — where they're going, what skills they're building, how fast they're leaving — the picture is a company in organizational decline, not one executing an AI transformation. Apple, with zero AI marketing, has 10-12% growth in Deep Learning and ML skills. Microsoft has 4% growth in Elasticsearch. That gap is significant.

For a megacap, the bleed is slower than it would be at a smaller company — but the direction is clear.