Cybersecurity company Sophos has expanded its partnership with OpenAI to bring frontier artificial intelligence models to managed service providers through Sophos Fusion, as the companies look to extend AI-powered security capabilities across the channel ecosystem.
The partnership will integrate OpenAI’s frontier models into Sophos Fusion, the company’s AI-native cybersecurity defence system. The platform connects security capabilities across endpoint, network, email, cloud, identity and security operations through a common architecture.
For managed service providers, or MSPs, the integration is intended to make frontier AI capabilities available within the security systems they already operate rather than requiring them to deploy and connect separate AI tools.
Sophos said partners will also be able to build recurring security services around the technology, potentially extending the use of AI across threat detection, investigation and response for their customers.
The announcement builds on an existing relationship between Sophos and OpenAI. In June, Sophos joined the OpenAI Daybreak Cyber Partner Program, an initiative focused on bringing advanced AI capabilities into defensive cybersecurity tools, services and workflows.
Through the Daybreak programme, Sophos has been adopting OpenAI capabilities in phases, initially focusing on defensive workflows with controls and human analysts remaining involved. Early areas identified by the company included accelerating managed detection and response investigations, strengthening security assessments and helping customers identify and remediate exposure.
The latest partnership places a more specific emphasis on the channel and MSP ecosystem.
Sophos operates a global partner network that includes managed service providers, managed security service providers, resellers and distributors. The company said the OpenAI integration is designed to make frontier AI available through this existing distribution and services structure.
The development comes as artificial intelligence increasingly affects both sides of cybersecurity. Advanced models can assist defenders in analysing threats and identifying vulnerabilities, but similar capabilities can also be used by attackers to accelerate vulnerability discovery and exploitation.
Sophos is also expanding its focus beyond using AI for cyber defence to securing the AI tools businesses themselves are adopting.
The company is making Sophos AI Defense available for early access in August 2026. The offering is designed to give organisations visibility into AI and so-called shadow AI usage across their environments, apply policies to that activity and help govern how sensitive data is used across AI models and applications.
Sophos said it plans to expand these protections with OpenAI to cover AI workloads, data and usage within its connected cybersecurity architecture.
The company already uses artificial intelligence within its security operations. Sophos says its managed detection and response operation resolves 52% of cases end to end using AI, with an average response time of 89 seconds, while maintaining human analyst oversight. These performance figures are company-reported.
The OpenAI partnership adds frontier model capabilities to this broader AI security strategy while retaining an emphasis on governed deployment.
For Sophos, the channel-focused approach provides a way to distribute advanced AI capabilities through MSPs that already manage cybersecurity for businesses. For partners, the integration could provide additional tools for developing AI-enabled security services without requiring them to independently integrate frontier models into their existing technology stacks.