Answer-First Lead
Coinbase laid off roughly 700 employees — 14% of its workforce — on May 5, 2026, and CEO Brian Armstrong made clear this isn’t just a headcount reduction. Teams are being flattened, middle managers replaced by AI systems, and some groups reduced to a single human overseeing AI agents. This is the second wave of AI layoffs: not efficiency theatre, but genuine org-chart rewiring.
🔍 THE BOTTOM LINE
AI isn’t just eliminating workers anymore — it’s eliminating the layers between workers. If your job is coordinating other humans who do the work, you’re in the crosshairs.
What Happened
Coinbase’s May 5 layoffs hit roughly 700 people across engineering, product, and operations. In an internal memo reported by the New York Times, Armstrong framed the cuts as part of a deliberate shift toward “AI-native operations.”
The key detail: entire management layers are being removed. Teams that previously had a manager, a senior manager, and a director are being compressed. Some teams are now a single individual contributor managing AI agents that handle what three middle managers used to do — scheduling, status tracking, code reviews, resource allocation.
This is different from the 92,000 tech jobs cut in 2026 so far. Those layoffs, while AI-attributed, were largely headcount reductions — doing the same work with fewer people. Coinbase is redesigning how work flows.
The Second Wave of AI Layoffs
The first wave — which we’ve covered extensively with Meta’s 8,000 cuts and Microsoft’s voluntary buyouts — was about cost savings. Companies used AI to justify doing more with fewer people. The work didn’t change; the staffing model did.
The second wave, exemplified by Coinbase, is structural. The work itself is being reorganised around what AI can do:
- Middle management compression — AI handles coordination, reporting, and basic decision-making. Fewer humans needed between leadership and execution.
- Agent-mediated teams — Instead of a manager overseeing 8 people, one person oversees 8 AI agents plus 2-3 human specialists.
- Flattened hierarchies — The “AI-native” org chart has fewer levels. Senior ICs report directly to leadership, with AI handling the administrative glue that middle managers provided.
What is an AI-native org chart? An organisational structure designed around AI agents handling coordination, reporting, and routine management tasks, reducing the need for human middle management. It works by assigning AI agents to handle scheduling, progress tracking, resource allocation, and basic approvals, while humans focus on strategic decisions and creative work. For example, Coinbase teams now operate with one human managing multiple AI agents instead of multiple human managers.
Why Middle Management Is the Target
Middle management has always been the corporate layer most vulnerable to automation — and not just because it’s expensive. Consider what most middle managers actually do:
- Status aggregation — Collecting updates from teams and reporting up. AI dashboards do this in real-time.
- Resource coordination — Allocating people to tasks. AI project management tools optimise this algorithmically.
- Quality gates — Reviewing work before it ships. AI code review, content review, and compliance checks are already deployed at scale.
- Process enforcement — Making sure people follow procedures. AI workflow systems enforce this without negotiation.
The parts of middle management that aren’t automatable — mentoring, strategic judgment, navigating office politics, making calls with incomplete information — are exactly the parts that senior leaders handle themselves. The middle was always coordination, and coordination is what AI does best.
As we noted in our piece on AI killing the entry-level career pipeline, the restructuring isn’t just hitting juniors. The Yale research showed compression happening at every level below senior leadership. Coinbase is the most explicit example yet of a company acting on that logic.
The Numbers Behind the Shift
| Metric | Value | Source |
|---|---|---|
| Coinbase employees cut | ~700 (14%) | New York Times |
| 2026 tech layoffs year-to-date | 92,000+ | Layoffs.fyi |
| AI-attributed layoffs (2026) | ~47.9% of total | Layoffs.fyi/CNBC |
| AI jobs eliminated per month (2025) | ~25,000 | Goldman Sachs |
| Middle management roles at risk globally | 30-40% | McKinsey (2025) |
What This Means for NZ
New Zealand’s tech sector is small enough that a single Coinbase-style restructuring at a major employer would be front-page news. But the real impact is indirect:
- Remote work pipeline shrinking — Kiwi developers who previously landed remote roles at US crypto/fintech companies are finding those positions disappearing or being “AI-augmented” into part-time contracts.
- Management career paths compressing — The traditional Kiwi career arc (junior → mid → senior → manager → director) assumes management layers exist. When companies flatten, you can’t climb a ladder that’s been dismantled.
- Local companies will follow — NZ’s fintech sector watches global players closely. If Coinbase’s AI-native model delivers results, expect Xero, TradeMe, and others to experiment with flattened structures within 12-18 months.
The AI automation engineer boom shows where the jobs are going. But for every new automation engineer role created, multiple management positions disappear. The net equation is brutal.
The Bigger Picture
Coinbase is a crypto company in a volatile market — you could dismiss this as industry-specific pain. But Armstrong’s language mirrors what we’re hearing across tech: AI isn’t a tool you use within your existing structure. It’s a reason to change the structure itself.
Gartner predicts 80% of organisations will reduce autonomous workforce levels by 2027. The Coinbase model — flattened, agent-mediated, AI-native — is what that looks like in practice.
The uncomfortable truth: if your entire job is moving information between humans, AI can do that faster, cheaper, and without needing a corner office. The humans who survive this wave are the ones who create information, not just relay it.
❓ Frequently Asked Questions
Q: What does this mean for NZ? NZ tech workers in remote roles at global companies are most at risk, especially those in coordination-heavy positions. Local companies will likely experiment with similar flattening within 12-18 months. Kiwis should focus on skills AI can’t replicate: strategic judgment, creative problem-solving, and direct revenue generation.
Q: Is this really different from normal layoffs? Yes. Traditional layoffs reduce headcount but keep the org structure. Coinbase is removing layers of the structure itself — replacing management with AI coordination. The company doesn’t just have fewer people; it has a fundamentally different shape.
Q: What should I do if I’m a middle manager? Start building skills that AI can’t replicate: strategic decision-making, stakeholder relationships, cross-functional leadership, and domain expertise. If your primary value is coordinating information flow, start transitioning now — the Coinbase model will spread.
🔍 THE BOTTOM LINE
Coinbase didn’t just cut 14% of its workforce — it showed what the AI-native company looks like. Fewer layers, fewer managers, more agents. The second wave of AI layoffs isn’t about doing the same work with fewer people. It’s about redesigning work itself. If you’re in the middle of an org chart, your seat just got a lot less secure.
SOURCES
- New York Times (May 5, 2026)
- Coinbase internal memo
- Layoffs.fyi
- Goldman Sachs
- McKinsey Global Institute (2025)
