Fastly, the edge cloud platform known for accelerating and securing web applications, closed Q4 2025 with record figures and a clear message to the market: Artificial Intelligence not only consumes more infrastructure but is also rewriting the traffic patterns flowing through the network.
The company reported revenues of $172.6 million for the quarter, a 23% year-over-year increase, and significantly improved its operational profitability. Meanwhile, its CEO, Charles “Kip” Compton, described AI as an increasing “tailwind” for Fastly’s business, especially due to the rise in traffic generated by agents and the need to manage bots and crawlers more precisely.
A financial turnaround the market is embracing
On a strict accounting basis, Fastly hit several milestones simultaneously: record quarterly revenue, margin improvements, and adjusted operating profit. The company reported its GAAP gross margin at 61.4% (up from 53.4% a year earlier) and its non-GAAP gross margin at 64.0%.
This “muscle” shift was also visible in the adjusted earnings per share: $0.12 (non-GAAP, diluted), compared to losses a year prior.
The immediate market response was clear: after releasing the results, shares reacted strongly with a significant after-hours jump, indicating investors see a pivot point in the company’s trajectory.
AI as a driver: more “machine-to-machine,” more pressure… and more opportunities
Beyond the quarter, what’s important is why Fastly believes AI is fueling its growth.
Compton explained that they are seeing more traffic related to agents (formerly described as “machine-to-machine”), because AI tools tend to query and verify more sites to respond or execute tasks. For Fastly, this means more requests passing through its network for customers who prioritize performance and availability.
The second vector is the AI workload at the edge: from training dataset processes to inference and related tasks requiring proximity to the user or low latency. The company views this as a natural evolution: as AI integrates into daily products and workflows, part of the computing and observability infrastructure is moving closer to the network’s edge.
And the third point is almost an inevitable consequence: as automated traffic grows, so does the need to distinguish between “good” bots and malicious bots. Fastly emphasizes that many clients want to “be present” in the AI world (allow certain crawlers or fetchers), but block those that degrade performance, increase costs, or pose risks.
This balance is not theoretical. In its Threat Insights Report, Fastly quantifies the importance of “desired” bots (crawlers and fetchers) and details the impact of AI crawlers: for example, noting that Meta and ChatGPT account for a significant share of AI crawler/fetcher traffic analyzed.
Key metrics: contracts, customers, and retention
In infrastructure businesses, the pulse isn’t measured solely by the quarter. Fastly closed with Remaining Performance Obligations (RPO) of $353.8 million, a 55% increase year-over-year, indicating a stronger base of committed revenues. It also increased its enterprise clients to 628 and its net retention over the last 12 months to 110%.
In other words: the thesis isn’t just “growth driven by AI,” but “growth with higher-quality customer portfolio,” with signs that existing clients are expanding their spending.
2026 guidance: the bar is raised
For 2026, Fastly projects revenues of $700 to $720 million and an non-GAAP operating profit of $50 to $60 million, with a non-GAAP EPS of $0.23 to $0.29. For the first quarter, it guided revenues of $168 to $174 million.
The implicit message is that the business is not only “sustaining” the internet’s transformation but aiming to capitalize on it: more automation, more agents, more AI, and consequently a greater need for edge computing, security, and traffic control.
Frequently Asked Questions (FAQ)
What is Fastly, and why is it considered an “edge cloud” platform?
Fastly operates distributed infrastructure to accelerate content delivery, run functions close to users (edge compute), and enhance security and observability for web applications.
Why does AI increase network traffic?
AI agents and tools tend to perform more automated checks and queries (crawling, fetching, source verification), which increases machine-to-machine requests compared to traditional human traffic.
What does “AI bot mitigation” mean, and why does it matter to companies and media?
It encompasses techniques to allow legitimate bots (like useful indexers or fetchers) while blocking malicious or abusive automation that could cause outages, fraud, aggressive scraping, or infrastructure costs.
What is RPO, and why is it frequently mentioned in results?
RPO (Remaining Performance Obligations) reflects committed revenues from existing contracts not yet recognized; it serves as a gauge of business visibility and future demand.
via: investors.fastly