The June 2, 2026 White House Executive Order Promoting Advanced Artificial Intelligence Innovations and Security (“AI Executive Order”) is a bold call to action for the federal government and critical infrastructure to work “collaboratively with the private sector to modernize government and private sector information systems and harden them against external threats.” JetStream Security stands ready to support the federal government in the execution of this executive order.  

“It’s come as no surprise to JetStream that the United States has to be best positioned to accelerate AI adoption and build public trust in its use by government, all while establishing itself as a leader in the AI space,” said Paul Loeffler, JetStream’s Director of Compliance & Partnerships. “But to build that house, you need a strong foundation. Ours starts with the federal architecture in mind: FedRAMP High, multi-tenancy, and designed for the enterprise of federated agencies.” 

Loeffler previously served for over a decade as the IT Program/Project Manager for the Cybersecurity and Infrastructure Security Agency (CISA)’s Continuous Diagnostics and Mitigation (CDM) program. CISA is a component of the U.S. Department of Homeland Security. 

The AI Executive Order calls for hardening Federal and critical infrastructure systems with AI-enabled defensive cybersecurity tools, then expanding access to those tools across agencies and infrastructure operators. It also calls on the U.S. Department of Homeland Security to release Binding Operational Directives to expedite and prioritize cybersecurity services that enhance AI-enabled defensive tools to counter the potential cyber risks in frontier AI models, and directs the Director of the Office of Management and Budget, in coordination with the National Cyber Director and the Director of CISA, to ensure Federal funding that can be directed toward applicants developing “advanced AI vulnerability detection.”  These tools will be critical for the next phase of the initiative, continuous monitoring of the frontier AI models once implemented in organizations after passing their initial safety screening.   

“For the longest time, systems were built to do a job, perform a function, connect to other systems, store data,” said Loeffler. “The federal workforce was constrained by what the system allowed. Those constraints are dissolving, and federal employees now have access to frontier models that aren’t relational to databases with strict user permissions. They’re agentic, ever-changing, and built to act. With one click, or just a screenshot handed to a model to process, an employee can expose intellectual property, CUI, or worse, all to be more productive. 

“JetStream is exactly that type of platform that can govern those frontier models that warrant additional oversight, after they clear the Federal government’s initial cyber safety screening and is deployed on actual information systems,” said Loeffler.  “It continuously monitors the AI models that carry elevated risk after deployment across enterprises and informs them where those frontier AI models are running, who is using them, and whether usage has drifted from approved configurations.  JetStream provides the monitoring and enforcement layer the initiative is counting on and expecting.” 

Once a “covered frontier model” is made use of within an organizations’ systems, JetStream not only provides continuous monitoring, but governs how it is used, by whom, and under what constraints, associating every interaction with an identity and a policy.  This operationalizes the trust framework that the AI Executive Order is intended to build. 

JetStream AI Hub™ inspects AI data flows across human, non-human and agentic usage and enforces governance policies in real time.  JetStream AI Drift Detection™ detects when AI systems deviate from approved configurations.  JetStream Verified MCP™ also provides customers with a secure MCP catalog, through which we provide security-hardened versions of MCP servers for deployment at the AI Hub. MCP servers, like other third-party, open-sourced software, can have vulnerabilities. Given its increased usage, it has become a vector for threat actors to actively exploit environments. The JetStream MCP Catalog provides a solution to this rapidly emerging challenge.   

JetStream AI Manifest™ and AI Blueprints™ provide the visibility and operational awareness a classification system like this demands downstream.  JetStream’s AI Manifest provides the first AI configuration management database (CMDB) which continuously catalogs and provides configuration details on AI Assets and usage across an environment. JetStream’s AI Blueprints provides insight into what is happening inside an AI workflow but also monitors and alerts if changes are made to the approved flow or agentic drift occurs.   

JetStream’s capabilities extend beyond those directly articulated in the AI Executive Order, enabling the enforcement of governance policies across active AI workflows using JetStream AI Blueprints™.  Customers can use JetStream’s Identity Broker™ and Key Broker™ to bind identity to every AI action regardless of whether the actor is human, machine or autonomous, even for unapproved shadow AI. 

When considering the AI Governance landscape, the JetStream team has seen a range of vendor approaches; some are focused on development activities only, others are focused on web usage, or some other singular component of AI adoption and usage seen within environments. JetStream Security is the only vendor to provide a complete end-to-end AI governance and security platform. JetStream’s Security-First AI Governance (SAIG) Platform focuses on the five AI trust gaps that most challenge AI adoption: Visibility, Design Control, Runtime Governance, Agentic Identity, and AI Accountability. Its capabilities enforcement extend reaches beyond those surfaces directly articulated in the where AI Executive runs: Order, local, enabling web, enforcement of the governance policies across multiple surfaces: local AI usage, web-based usage, cloud usage, cloud, and SaaS tools. “We pull all of that together in the JetStream SAIG Platform™, which consolidates visibility, identity, governance and enforcement into a single monitoring, identity and governance layer,” said Loeffler.  “This provides the unified command structure necessary to fully see and govern covered AI models in your native environment, this is the operational control layer that enables enterprises to operationalize policy and bridge the gap between model-level risk and real-world usage.” 

The June 2 order started a 90-day clock. JetStream helps federal teams act on it with a plan built for their seat. Book a consultation and we will map your 90-day AI action plan together.