The Exchange Daily

HHS restores unified OCIO oversight of AI, data, and tech roles while GAO, NIST, and OPM deliver actionable guidance for federal AI programs. Plus: today’s Exchange Weekly delivers the complete executive guide to GSA’s proposed AI Procurement Clause GSAR 552.239-7001, detailing exactly what the new Basic Safeguarding of Artificial Intelligence Systems rules mean for federal contractors, agencies,...

Trump Administration Releases National AI Policy Framework Calling for Federal Preemption The White House blueprint outlines seven pillars and pushes Congress to establish one national standard while preempting unduly burdensome state AI laws. Agencies and contractors now have clear marching orders on innovation, liability, and governance. Data Readiness Emerges as the Top Barrier to Federal AI Sc...

California’s Executive Order Tightens AI Procurement Guardrails Governor Newsom’s March 30 order directs agencies to vet AI firms on exploitation prevention, civil-rights protections, and child-safety measures, creating a potential model for federal procurement teams navigating preemption debates. GSA and NIST Formalize AI Evaluation Partnership for Federal Procurement The new memorandum of unders...

White House National AI Legislative Framework Gains Traction on Capitol Hill The White House National AI Legislative Framework continues to shape congressional drafting as lawmakers work toward unified federal rules. Priorities include preemption of conflicting state laws, free-speech protections, streamlined data-center permitting, and small-business AI support. Federal IT leaders and contractors...

White House Releases National AI Legislative Framework The White House issued legislative recommendations for Congress that prioritize federal preemption of state AI laws, free-speech protections, streamlined data-center permitting, anti-scam measures, and small-business support. The framework advances American AI leadership with a light-touch approach. Federal IT leaders and contractors must alig...

White House Appoints Tech CEOs to President’s Council of Advisors on Science and Technology President Trump announced PCAST members including co-chairs David Sacks and Michael Kratsios plus leaders such as Jensen Huang (Nvidia), Mark Zuckerberg (Meta), Larry Ellison (Oracle), and Lisa Su (AMD). The council will advise on strengthening U.S. leadership in science and technology with emphasis on emer...

Treasury Launches AI Innovation Series to Strengthen Financial System Resilience The U.S. Treasury Department and FSOC launched the AI Innovation Series, a public-private effort to accelerate responsible AI use across fraud detection, cybersecurity, credit underwriting, and risk management. Deputy Assistant Secretary Christina Skinner emphasized that AI adoption is now essential for financial stab...

CISA Adds Five New Entries to Known Exploited Vulnerabilities Catalog The Cybersecurity and Infrastructure Security Agency has added five new vulnerabilities to its Known Exploited Vulnerabilities catalog, each backed by evidence of active exploitation in the wild. One critical entry is CVE-2025-31277, a buffer overflow impacting Apple Safari, iOS, macOS, and related products that could enable mem...

NIST Releases Report on Challenges Monitoring Deployed AI Systems NIST published AI 800-4 this month, the first federal report mapping gaps in post-deployment AI monitoring after workshops with over 200 experts and an 87-paper literature review. The document organizes monitoring into six categories and details barriers including performance drift, security risks, and limited data-sharing infrastru...

Marvell Launches 260-Lane PCIe 6.0 Switch for AI Data Centers Marvell Technology has introduced the Structera S series, establishing the industry’s first 260-lane PCIe 6.0 switch optimized for the demanding requirements of modern AI data center infrastructure. This breakthrough delivers approximately double the lane density available from previous leading solutions, allowing organizations to const...

White House Releases 2026 Cyber Strategy The administration’s new framework prioritizes zero-trust, AI defenses, and post-quantum readiness across federal networks. Agencies must harden systems aggressively amid procurement reforms. NIST Report on Monitoring Deployed AI Systems New guidance highlights real-world oversight challenges for production AI, urging continuous auditing to catch drift and...

Executive Summary The single most consequential development for system integrators and service providers this week was the Department of War’s formal designation of Anthropic as a supply-chain risk. Effective immediately after the March 4 notification, the action forces every contractor supporting War Department work to reassess and, where required, eliminate direct use of Claude models in any con...

CISA Issues Emergency Directive 26-03 to Mitigate Active Exploitation of Cisco SD-WAN Systems Federal agencies face an immediate 48-hour window to secure Cisco SD-WAN infrastructure under active exploitation. Inventory by midnight tonight and full remediation by 5 p.m. tomorrow are required. House Passes NIST Small Business Artificial Intelligence Advancement Act Lawmakers approved legislation dir...

AI Startup Funding Hits Record Pace: 17 US Companies Raise $100M+ in First Six Weeks of 2026 Seventeen US-based AI companies have each secured one hundred million dollars or more since January first. The surge includes Anthropic’s thirty billion dollar round at a three hundred eighty billion dollar valuation, Runway’s three hundred fifteen million dollar video push, and SkildAI’s one point four bi...

The Exchange Daily will soon be going to a subscription-only model. Don’t worry though, it will still be free. Subscribe now to avoid any disruptions. Thanks for reading The Exchange Daily! Subscribe for free to receive new posts and support my work. OpenAI $100B Funding, Pentagon-Anthropic Clash, CISA KEV Update, OMB Policy Reset, Federal AI ROI: The Exchange Daily Executive intelligence on the c...

DOE Genesis Mission and the 26 Grand Challenges The Department of Energy has formalized its Genesis Mission Consortium to accelerate scientific discovery. Alongside this, the agency announced 26 Grand Challenges aimed at doubling U.S. research and development productivity by 2036. Key focus areas include AI-driven power grid planning and the development of autonomous laboratory discovery systems t...

BeyondTrust Zero-Day Deadlines and the $650B AI Infrastructure Surge CISA mandates an immediate patch for BeyondTrust tools, Big Tech signals a massive 2026 capex surge, and the FBI launches Operation Winter SHIELD to harden infrastructure. BeyondTrust Critical Zero-Day (CVE-2026-1731) CISA has added a critical OS command injection vulnerability in BeyondTrust Remote Support and Privileged Remote...

CISA BOD 26-02: Federal Edge Device Purge CISA has issued Binding Operational Directive 26-02, requiring all Federal Civilian Executive Branch agencies to remove end-of-support edge devices. Agencies must complete an inventory by May 5, 2026, and decommission unsupported hardware within 18 months to mitigate risks from nation-state exploitation. DHS Shutdown and Federal Impact The Department of Ho...

OMB Memo M-26-04 Implementation Deadline Federal agencies must finalize procurement updates by March 11, 2026, to enforce “ideological neutrality” and transparency in all procured Large Language Models. This memorandum, which implements Executive Order 14319, makes these principles material to contract eligibility. Vendors must prepare to disclose detailed documentation regarding model training an...

OMB M-26-04: The End of “Black Box” Procurement Agencies have until March 11, 2026, to align with OMB Memo M-26-04. The directive enforces two “Unbiased AI Principles”: truth-seeking and ideological neutrality. This shifts the burden of proof to vendors, who must now disclose training methodologies and support independent evaluation of model behavior. The Genesis Mission Consortium The DOE is cent...

CISA Issues BOD 26-02 for Edge Devices The federal government is moving aggressively to secure the network edge. CISA’s latest Binding Operational Directive requires agencies to move beyond simple patching toward structural hardening and, in some cases, replacement of edge hardware that cannot meet modern security standards. The $80 Billion Sovereign Cloud Market Gartner’s latest data shows a mass...

Google Launches Vertex AI Agent Builder to General Availability Enterprises now have access to a production-ready environment for building and deploying AI agents. The release emphasizes data grounding, allowing agents to pull from private corporate repositories with high accuracy. For technical leadership, this provides the necessary monitoring and versioning tools required for mission-critical d...

CISA Adds Ivanti EPMM to KEV Catalog CISA has added CVE-2026-1281 to its Known Exploited Vulnerabilities catalog. This code injection vulnerability in Ivanti Endpoint Manager Mobile (EPMM) allows unauthenticated remote code execution. Federal agencies are mandated to remediate this vulnerability by February 1, 2026, signaling extreme urgency for the private sector. Microsoft Reports “Trillion Toke...

Google Cloud: Vertex AI Agent Builder billing shift and Agent Designer support for older agents Google updated Vertex AI Agent Builder documentation to move the charging start date for Sessions, Memory Bank, and Code Execution to February 11, 2026. That timing shift gives platform and FinOps teams a short runway to tighten cost controls, tagging, and budget guardrails before spend begins. Google a...

Google Cloud Vertex AI Agent Engine Billing Begins Feb 11 Google has updated the timeline for Vertex AI Agent Engine’s move to metered service. Billing for Sessions, Memory Bank, and Code Execution is now scheduled to begin February 11, 2026. This change requires FinOps teams to audit existing AI pilot projects and set spending limits before the new charges take effect next month. Sources: https:/...