ReliaQuest deepens Anthropic deal to bolster GreyMatter

ReliaQuest has expanded its collaboration with Anthropic by more deeply integrating Claude models into its GreyMatter cybersecurity platform, building on the companies' existing work on AI in security operations.

Under the arrangement, ReliaQuest is bringing Anthropic's core Claude model family into GreyMatter, the platform it sells to large organisations to detect, investigate and respond to cyber threats. It is also joining Anthropic's Project Glasswing and applying Mythos to its internal systems for defensive cybersecurity work.

The move broadens a relationship that already included Claude in parts of GreyMatter. The platform has used Claude for alert triage, threat investigation and the analysis of complex, multi-stage attacks.

It also comes as cybersecurity groups and their customers assess how generative AI is changing both attack methods and defensive tools. Across the sector, companies are placing greater emphasis on automation and AI-assisted analysis as security teams face growing alert volumes and increasingly sophisticated threats.

Brian Murphy, Founder and Chief Executive Officer of ReliaQuest, said the change reflects a wider shift in the threat landscape.

"We are at a defining moment for cybersecurity," said Brian Murphy, Founder and Chief Executive Officer of ReliaQuest. "AI is changing what attackers can do. It has to change what defenders can do too. Our work with Anthropic means our customers will have more access to frontier AI models deeply integrated into GreyMatter - so they can move faster and respond with greater confidence than the adversary."

Broader integration

Claude has been part of GreyMatter from the outset as one of more than 20 AI models used within the platform. Those models sit behind what ReliaQuest calls an AI Model Broker, which selects a model for each task based on factors including speed, accuracy and cost.

The deeper integration with Anthropic expands that use while also tying ReliaQuest more closely to Anthropic's efforts to examine the security risks around advanced AI systems. By joining Project Glasswing, ReliaQuest expects to gain more insight into how threat actors may use emerging AI models.

That defensive focus is also reflected in the company's adoption of Mythos in its own systems. ReliaQuest described that work as part of its internal defensive cybersecurity activity rather than a customer-facing product change.

Compliance monitoring

The companies have already worked together on compliance and oversight functions. ReliaQuest was among an early group of security vendors selected to integrate with Anthropic's compliance API, which brings authorised Claude activity data into GreyMatter.

That allows enterprise security teams using GreyMatter to monitor, detect and respond to Claude usage in much the same way they handle other enterprise applications in their environment. In practice, AI tool activity can be pulled into the same workflows already used for investigation and response.

The integration reflects growing concern among large organisations over how staff use external AI services and how those tools fit within existing security controls. As businesses roll out generative AI more widely, security teams are under pressure to track usage, ensure authorised access and spot anomalous behaviour.

Enterprise focus

ReliaQuest positions GreyMatter as a platform for large enterprises managing security operations across a wide range of technologies. The company says users can work across their stack in plain language, while the platform handles the underlying translation of data from different vendors and systems.

Its description of the product centres on three areas: normalising telemetry across different tools, detecting threats where data resides or while it is moving, and orchestrating automated actions through multiple AI systems. ReliaQuest argues that this structure helps security teams work faster without needing specialist knowledge of every tool in their environment.

Anthropic has been expanding the use of Claude models in fields including code work and security analysis. The deeper relationship with ReliaQuest shows how model developers are seeking more direct roles inside enterprise software rather than acting only as standalone AI providers.

For cybersecurity vendors, these tie-ups may also become a way to differentiate products in a crowded market, particularly as customers demand clearer evidence that AI features are tied to practical security tasks rather than broad claims about automation. ReliaQuest said GreyMatter customers will gain wider access to Anthropic's latest models through the expanded collaboration.