Cloudways Copilot adds AI diagnostics with human approval

Managed cloud hosting provider Cloudways has introduced Copilot, an AI-driven diagnostics tool designed to detect infrastructure issues and propose remediation steps—without automatically applying fixes. The tool, currently in preview, targets server-level problems such as DDoS patterns, disk space constraints, and database query inefficiencies, surfacing root causes within minutes. However, its design deliberately keeps humans in the loop, requiring explicit user approval before any corrective action is taken.

Copilot’s rollout reflects a broader industry push toward self-healing infrastructure, though Cloudways has opted for a cautious approach. The tool’s SmartFix feature offers one-click remediation for common issues, but adoption remains gradual, with customers often using it first for visibility before trusting it for proactive fixes. This hesitation underscores a key challenge: balancing automation with the risks of unintended consequences in production environments.

How Copilot works

Copilot analyzes real-time system signals to identify issues and recommend solutions, but its workflow separates diagnosis from execution. For example, if the tool detects a bot-driven traffic spike, it will flag the pattern and suggest mitigation steps—such as rate-limiting or IP blocking—but wait for user confirmation before implementing changes. This design choice stems from concerns about automated fixes disrupting live applications, particularly in complex or customized setups.

The tool’s current focus is on server-level diagnostics, with plans to expand into application-layer issues in future releases. Early data shows Copilot reliably identifies high-volume, pattern-based problems, though highly customized application issues still require human intervention. During the preview phase, nearly 400 customers tested the tool across diverse environments, validating its performance at scale.

Adoption and business impact

Copilot’s adoption has been steady but measured. Customers initially use it for visibility, gradually shifting toward proactive diagnostics as trust builds. The tool’s credit-based pricing—$9.99 or $19.99 per month—positions it as a paid add-on, reflecting its value in reducing operational overhead rather than bundling it into base hosting costs. This model aligns with usage patterns, as customers tend to engage with Copilot during periods of high activity rather than as a constant monitoring tool.

For agencies, Copilot’s appeal lies in its potential to streamline client-server management. While the tool currently operates on a per-application basis, Cloudways plans to introduce features allowing agencies to monitor and manage multiple client environments from a single interface. This could reduce the time spent troubleshooting individual issues, though an agency-specific pricing model has not yet been announced.

The tool’s impact on support costs is still emerging. While Copilot has reduced diagnosis time for common issues, its effect on Cloudways’ support team size or expenses remains unquantified. Hostinger’s AI assistant Kodee, by comparison, handled 81% of support interactions in 2025, saving the company approximately €9 million. Cloudways has not released equivalent metrics, though the company notes that Copilot’s gradual adoption suggests a longer path to similar efficiencies.

Long-term direction

Cloudways’ approach to AI diagnostics contrasts with competitors like Bluehost and WordPress.com, which focus on AI-driven site building and content generation. Copilot’s emphasis on infrastructure aligns with Cloudways’ core strength in managed hosting, though the company’s roadmap includes potential expansions into application-layer intelligence and even AI-assisted site setup. For now, the priority remains on improving production reliability and reducing operational friction for developers and agencies.

The tool’s human-in-the-loop design reflects a deliberate trade-off between automation and control. While the long-term vision includes more proactive remediation, Cloudways has ruled out fully autonomous fixes, citing the risks of unintended consequences in live environments. Instead, the company envisions a system where AI handles the heavy lifting of diagnosis and recommendation, while users retain final authority over execution.