Cloudflare is cutting more than 1,100 staff, about 20% of its global workforce, in what its founders are calling a shift to an “agentic AI-first operating model”. Chief executive Matthew Prince and co-founder Michelle Zatlyn announced the decision on 7 May 2026 in a joint blog post that ran alongside the company’s first-quarter earnings.
Prince and Zatlyn say this is an architecture change, not a budget cut. Internal AI usage at Cloudflare has grown more than 600% in three months, and the leaders are “reimagining every internal process, team, and role” around it. For Australian firms running on Cloudflare’s edge network, or quietly weighing similar moves of their own, it lands as one of the year’s clearer signals that agentic AI is reshaping headcount and not just tooling.
Inside the Cloudflare announcement
In a joint blog post titled “Building for the future”, the founders wrote: “We’ve made the decision to reduce Cloudflare’s workforce by more than 1,100 employees globally.” They were careful to add that the move was “not a cost-cutting exercise or an assessment of individuals’ performance”. According to an SEC filing, the restructure will cost the company about $150 million.
By tech-industry standards, the severance is generous. Departing staff keep their full base pay through to the end of 2026, US-based employees retain healthcare cover for the year, and equity vesting is brought forward to 15 August. None of which seemed to soothe investors. Despite first-quarter revenue of $639.8 million (up 34% on the prior year) and guidance of $664 million for the second quarter, Cloudflare shares dropped roughly 13% in after-hours trade.
The ‘agentic AI’ rationale

Cloudflare’s argument is that the way it works internally has changed faster than its headcount could keep up with. Internal AI usage has grown more than 600% in three months, the company says, with employees running “thousands of AI agent sessions daily”. Prince told The Register that “the people embracing these tools are so much more productive than we have ever seen before”.
He also clarified, after the announcement, that “very few engineers or customer-facing sales people” were affected. The cuts landed largely on back-office and internal-process roles: procurement, IT support, knowledge management and the like. The distinction is worth noting. Cloudflare isn’t claiming AI can replace customers’ favourite humans. It’s saying AI can absorb much of the administrative scaffolding the company runs on.
Why Australian businesses should care
Australian organisations aren’t bystanders here. Several of the country’s biggest banks, retailers and government services run on Cloudflare’s edge network and security products, which makes a restructure of this size a fair moment to ask the obvious continuity questions. Who’ll be answering support queries. How SLAs hold up. Which features get prioritised while teams reform around AI agents.
The wider context is harder to ignore too. The global tech sector logged close to 99,000 layoffs in the first months of 2026, and Cloudflare’s framing, that AI is changing the operating model rather than the cost base, is the line Australian boards will hear more often this year, including from local outfits looking for cover to do the same thing. The honest test is whether the productivity gains actually show up, or whether the cuts simply outrun the agents.
What it signals for AU automation strategy
For SMEs and mid-market businesses across Australia, the practical question isn’t whether Cloudflare got it right. It’s what to take from any of it. The signal is that back-office automation, procurement, knowledge bases, IT triage and document review, is where agentic AI is delivering the clearest productivity gain right now. Customer-facing roles paired with conversational AI assistants look more like augmentation than replacement.
That distinction is worth keeping in mind. The businesses that come through this period well are the ones pairing AI agents with redesigned processes, not the ones counting seats and calling it a strategy. As Prince and Zatlyn put it, “reimagining every internal process, team, and role” is the hard part. Announcing layoffs is the easy bit.
Watching the next quarter
Cloudflare’s restructure is the largest public test so far of a deliberately AI-first operating model from an infrastructure-adjacent company. The next two quarters should reveal whether the productivity claim holds, measured against revenue growth, customer churn and product velocity, or whether the cuts simply ran ahead of the agents.
For Australian leaders, the takeaway is closer to careful than dramatic. Working out which roles in your own organisation can genuinely be augmented by AI agents, and which still rest on human judgment and customer trust, is the kind of question Cloudflare’s announcement makes a bit harder to put off.
Done well, AI agents free people up for the work that actually compounds: relationships, design, judgment, the awkward edge cases. Done carelessly, you end up with an organisation that’s leaner and slower at the same time. Cloudflare has placed its bet. The rest of the market is watching.
