Cloudflare's fourth-quarter results saw the web giant win from demand for distributed networking and edge services, as well as extol the virtues of the growing traffic bugbear that is agentic AI.

The firm's Q4 revenue totaled $614.5 million at an increase of 34% year-over-year (YoY), with fiscal year 2025 revenue of $2.17 billion, representing an increase of 30% YoY.

Co-founder and CEO Matthew Prince flagged significant deals in the quarter, including Cloudflare's largest annual contract value deal ever, averaging $42.5 million per year. The exec touted Cloudflare being able "to fill critical gaps in the hyperscaler model" by tempting the likes of a Fortune 500 technology company which found such providers were unable to handle the complexity of global third-party applications at scale.

A major international consumer goods company, meanwhile, did away with the "fragmented architecture of multiple legacy vendors," which led to significant operational complexity and security gaps, to extend its Cloudflare contract and leverage features such as its serverless Workers platform that allows users to run code on the edge via Cloudflare's global network.

"Despite an incumbent offering significant discounts to keep the business, this customer chose Cloudflare because of our next-generation architecture and Workers developer platform, which allowed them to customize their edge traffic," Prince claimed. "We didn’t just replace three incumbents. We replaced a legacy mindset, proving once again that when you give developers the best tools, the entire enterprise follows."

Agentic flair

Cloudflare Workers was also mentioned in regard to AI and agents driving more demand for Cloudflare services. Prince cited a U.S. media company that signed a three-year, $3.1 million contract for Workers, along with Cloudflare's AI scraping prevention tool and various application services. Echoing Cisco's anticipation last year of an "80 billion" bandwidth blitz from numerous AI agents now circling the web, the customer was facing a "massive increase" in scraping that was impacting their network and driving up infrastructure costs.

"They chose Cloudflare to gain visibility into which AI models are consuming their data, allowing them to protect and eventually monetize their unique content," Prince said. "By leveraging Cloudflare Workers to replace years of complex technical debt from an incumbent, they were able to migrate massive internet properties into production in just two weeks."

In Prince's view, the deal proved that as AI accelerates, Cloudflare is the "partner of choice for companies looking to protect their IP while improving performance, reducing operational costs, and enhancing their security postures."

The CEO also talked up AI driving demand for Cloudflare's Zero Trust platform.

"AI usage is even driving adoption of our Zero Trust platform to ensure that data is compartmentalized and access granted in limited and controlled ways," Prince told investors. "[It's] getting more and more adoption because it’s built from a sense of agentic AI from the beginning. [And] our standard application services platforms are getting more adoption because people need to be able to keep up with all this new traffic that’s coming to the internet."

Cloudflare itself is contributing to the AI wave online, having recently acquired AI deployment startup Replicate, which focuses on the streamlined creation of AI models and agents.

Its Workers toolkit, meanwhile, includes agents like Moltworker, which enables agentic tool du jour OpenClaw to run on Cloudflare's Sandbox SDK and developer platform APIs.

Previously known as Moltbot, OpenClaw's system lets AI agents execute, coordinate, and delegate tasks to other agents automatically – much to the concern of security guardians like Palo Alto Networks, who warned interconnection between AI agents isn’t necessarily secure and that such a level of ungoverned autonomy can give rise to “irreversible security incidents.”

Besides OpenClaw, a forum run and used exclusively by agents by the name of Moltbook has also caused some cause for alarm. SDxCentral last week discovered AI agents, some of whom operate on the OpenClaw system, sharing security evasion tips on the so-called Reddit for AI agents.

In response to the report, Christian Reilly, field CTO EMEA at Cloudflare, said the emergence of a closed-loop forum where AI agents are able to swap scraping tactics is “more of a mechanized evolution of the existing arms race.”

“While it feels like the ultimate normalization of evasion, it represents the ‘added AI sheen’ phase where the volume and velocity of bot-to-bot knowledge sharing outpace human intervention," the CTO said. "[This is] the transition from static fingerprinting to defending against an ecosystem where bots autonomously learn to bypass TLS fingerprints and solve CAPTCHAs by consulting a communal database of failures and successes.”

“We should not be worried about the agents liking each other's posts,” Reilly added, “but about the security surface area. Cybersecurity researchers have already found that Moltbook’s lack of rate-limiting and its Skills framework make it a prime vector for indirect prompt injection and RCE (remote code execution). If your agent learns a scraping tip from a malicious agent on Moltbook, it might actually be downloading a script that exfiltrates your API keys.”