Bedrock Data MCP Server Now Generally Available for Governing Agentic Workflows

Business Wire

6 days ago

MENLO PARK, Calif. & LAS VEGAS--(BUSINESS WIRE)-- Bedrock Data (formerly Bedrock Security), the platform provider for data-centric security, governance and management, today announced the general availability (GA) of its Model Context Protocol (MCP) Server. The MCP server is an essential component in protecting agentic AI systems and harnessing agentic AI for enhanced security operations. It provides two key capabilities for enterprises: it enables agentic AI systems to self-govern by drawing on Bedrock's Metadata Lake to provide data context for autonomous decision-making and governance; it also allows security teams to leverage AI-driven metadata intelligence for more precise, faster incident response through alert prioritization and context-aware threat detection. As part of Bedrock's growing strategic partnership ecosystem, the company also announced an integration with Panther, the security monitoring platform for the cloud, demonstrating how the MCP Server's capabilities can enrich real-time signals with data risk context to improve fidelity and prioritization of alerts, reduce false positives, and improve the security operations centers' (SOC) response to incidents.
'The MCP Server, built on Bedrock's Metadata Lake, gives enterprises a way to govern how AI agents and operators access data – enabling safe, controlled deployment of agents across their environments,' said Pranava Adduri, CTO and co-founder of Bedrock Data.
The Enterprise Standard for AI Agent Governance
The GA of the Bedrock Data MCP Server enables agentic workflows to incorporate data context from Bedrock's Metadata Lake, including classifications, lineage, entitlements, business taxonomy and more, providing agents with governance context needed to handle sensitive data correctly. It allows self-governing policy enforcement and governance controls throughout the agent lifecycle, from initial authorization through ongoing execution, to automatically adapt to evolving data classifications, location changes and policy updates. For example, if an organization prohibits agents from using customer data, by introducing context from the Metadata Lake on the data being handled, agents can now be aware of sensitive data and refuse to process it according to company policy. This approach delivers more efficient agentic infrastructure by eliminating the need for additional governance agents.
'MCP servers are rapidly gaining popularity for enabling communication between generative AI applications and the data they utilize. The integrity and availability of the models these MCP servers support rely on proper classification and governance of sensitive data,' said Krista Case, research director of cybersecurity at The Futurum Group. 'Additionally, the ability to harness metadata for context-aware alert prioritization and threat intelligence will be needed in order to thwart data breaches. Bedrock Data's new MCP Server offering represents an operational solution squarely addressing this emerging and increasingly critical requirement.'
Bedrock Data and Panther Collaboration Transforms Security Response With Context-Aware AI
With the Bedrock MCP Server, enterprises can integrate into other MCP clients to integrate data risk context into their workflows such as Splunk (SIEM), Vanta (compliance), Wiz (CNAPP) and Okta (identity), enhancing the security and governance of their entire AI ecosystem. The integration with Panther is the newest addition to Bedrock's growing security and data partner ecosystem.
'Through our partnership with Panther, we're demonstrating how AI-powered security tools can dramatically improve response times and alert prioritization while maintaining proper governance controls and data context,' continued Adduri. 'It puts data, the crown-jewel of digital enterprise, at the center of the security process.'
Powered by their respective MCP Servers, the Bedrock and Panther integration improves SOC team effectiveness by enabling AI tools used in triage to leverage contextual alert prioritization and reduce false positives. Each Panther alert is enriched with Bedrock's data context, including sensitivity classifications, storage locations, identity risk profiles and regulatory mappings, helping security teams respond with greater precision and speed.
The partnership delivers key capabilities beyond traditional webhook integrations, including:
AI-driven investigation agents: Security teams can leverage Claude, Rewind and other AI assistants to enable enriched security context without leaving their interface, dramatically reducing mean time to investigate (MTTI). They can ask questions like 'Does this alert involve regulated PII?' or 'Does this identity have access to production customer data?' directly within their AI interface.
Data-aware alert correlation: Bedrock enriches Panther alerts with real-time sensitivity, access and usage context, allowing correlation rules to factor in actual data risk, not just event volume or signature matches.
Improved signal-to-noise: Alerts are prioritized based on what matters most: whether sensitive or regulated data was involved, and whether high-risk identities accessed it. This reduces false positives and surfaces the alerts that deserve immediate attention.
Faster root cause analysis: Analysts can quickly determine what data was at risk, how it moved and who accessed it, without pivoting between systems. This accelerates investigations and shortens containment time.
AI-enabled investigation workflows: Through Bedrock's Copilot and metadata API, analysts and AI assistants can ask structured questions like 'Did this identity access PHI this week?' directly within their workflow, improving investigation quality and speed.
'The collaboration with Bedrock represents a fundamental shift in detection and response,' said Jack Naglieri, founder and CTO of Panther. 'By utilizing AI agents to reach into Panther's analysis capabilities and Bedrock's data context, security teams can make informed decisions in half the time. A natural language interface also opens these strong capabilities to everyone in the security team. When every alert comes with built-in understanding of data sensitivity, regulatory impact and blast radius, analysts can make confident decisions in seconds rather than hours.'
Examples of how customers can implement the Bedrock-Panther integration to improve their security operations include:
RDS without KMS: Panther flags an unencrypted RDS instance and Bedrock identifies it as holding production customer PII. This prioritizes a sensitive RDS instance for remediation over others, such as test/QA instances that don't hold sensitive data.
IAM user without MFA: Panther detects privilege escalation and Bedrock shows the user has broad privileges to access regulated HR data. It's not just a misconfiguration; it's a data exposure risk.
Suspicious privilege escalation: Panther flags an unexpected change in access rights. Bedrock adds visibility into what data the account accessed after the change, including whether any of it was sensitive. What used to be a vague alert becomes a focused, actionable response based on real data exposure.
Availability
The MCP Server is available on GitHub or as a self-hosted offering.
Learn how your team can get started with the MCP Server and the Bedrock-Panther integration today.
Additional Resources
About Bedrock Data
Bedrock Data equips enterprises to secure and manage all of their data with confidence. Powered by patented Metadata Lake technology and AI-driven classification, the Bedrock platform provides continuous visibility into data location, sensitivity, business context, access and usage across SaaS, PaaS, IaaS and agentic AI workflows. By autonomously discovering, classifying and contextualizing all enterprise data, Bedrock Data empowers security, governance and data teams to tackle security, roll out AI technologies with confidence and manage their data at multi-petabyte scale and minimal infrastructure cost. Bedrock Data's open, API-first architecture, built on its Metadata Lake, powers its built-in MCP Server to deliver real-time metadata context to agentic AI workflows, driving accurate, governed AI decisions. This metadata context also enriches tools like SIEM, DLP, CNAPP and cloud platforms to amplify visibility, risk insights and response precision across the enterprise. Trusted by leading enterprises worldwide in financial services, healthcare, biotech and other industries, Bedrock Data is headquartered in Silicon Valley, backed by Greylock Partners, and led by experts in cloud infrastructure, generative AI, cybersecurity and large-scale data management. Learn more at bedrockdata.ai.
About Panther
Panther is the security monitoring platform for the cloud. Unlike 'next-gen' SIEMs that rely on historical detections, rigid query languages, and closed ecosystems, Panther enables flexible monitoring in production environments with streaming data analysis, programmable workflows, and seamless cloud integration – empowering teams to optimize costs and control, accelerate incident response, and achieve cross-system visibility at scale.