Nine thousand, two hundred and thirty-eight. That's how many of this year's 45,363 tech layoffs are explicitly attributed to AI restructuring — and that only counts the cases where companies said so out loud. Five days of news this week suggest the real number is considerably higher.

---

⚙️ Atlassian Cuts 1,600 — and Its CTO — for AI
Enterprise SaaS / Workforce Economics

Atlassian shed 10% of its global workforce on March 11, replacing its CTO simultaneously with two executives described as "next generation AI talent." CEO Mike Cannon-Brookes called it "the right decision." No qualifiers.

The reason Atlassian matters more than Meta here isn't the numbers — it's the precedent. Meta is a megacap absorbing regulatory pressure and hedging superintelligence bets. Atlassian is a mature, profitable, 17,000-person B2B SaaS company that just ran the same trade: cut headcount, fund AI reinvention, install AI-native leadership. That template is now available to every enterprise software shop below them. When the mid-market runs it — and they will — it won't make the same headlines, but the aggregate will be larger.

---

📉 Meta: $600B In, 20%+ Out
Big Tech / AI Infrastructure

Meta committed $600B to AI data centers this year. Nine-figure pay packages went out to recruit a superintelligence research team. And 20%+ of the human workforce is going out the door to offset the cost.

The contrast isn't ironic — it's the arithmetic. The same company simultaneously laying off workers and hiring AI researchers isn't confused; it's executing a specific bet about where value is created next. The workers funding that bet are the ones who built everything Meta has. CEOs have historically been careful about saying that quiet part out loud. This week, they stopped.

---

📊 ServiceNow CEO: 30%+ Unemployment for Recent Grads, Easily
AI Agents / Labor Economics

ServiceNow CEO Bill McDermott told CNBC that AI agent adoption could push unemployment for recent college graduates "easily into the mid-30s in the next couple of years." He wasn't speculating — he was forecasting from data. NY Fed numbers put underemployment for new grads at 42.5% at end of 2025, the highest since 2020. McDermott's 30%+ unemployment projection is a step function from that baseline, not a jump from nowhere.

CEOs don't say this on CNBC unless they've made peace with the implication. McDermott is describing the endpoint of the Atlassian and Meta trade — run forward a few years, applied across the software economy. If you build software for a living, or you're training someone who will: this is a CEO speaking, with data, on the record. That's a different weight than a researcher's paper.

---

The pattern across all five days: companies are no longer hedging the trade-off. They're announcing it. The AI displacement thesis was always going to arrive. The tell that it has is when the people making it stop using euphemisms.

📡 More signal, less noise → www.thesignal.press