THE APEX TIMES
Survey by Booz Allen Finds Federal Agencies Accelerating Agentic AI, but Security and Trust Lag
More than half of federal IT and cybersecurity decision makers say their agencies are already deploying or piloting AI systems that can act autonomously, but fewer than one-third express high confidence they can do so securely, according to a new Booz Allen survey.
Federal agencies are moving quickly toward “agentic” artificial intelligence, systems designed not just to generate responses but to take actions, yet a sizable trust and security gap remains, according to a new survey released by Booz Allen Hamilton.
The survey, published July 21, is featured in the latest issue of Velocity, Booz Allen’s ongoing publication focused on emerging technology and mission innovation. It examines what government cybersecurity leaders see as opportunities and risks as autonomous AI enters IT and cyber operations.
Among more than 100 federal IT and cybersecurity decision makers surveyed, 58% said their agencies have already deployed or are piloting AI agents. However, only 28% said they have high confidence in their ability to deploy those systems securely, suggesting a widening divide between adoption plans and governance readiness.
Booz Allen said the survey also points to continued concerns about adversaries using AI-enabled cyber capabilities. Seventy-nine percent of respondents reported they are “very” or “extremely concerned” that opponents could use AI to accelerate cyberattacks against their agencies over the next 12 to 18 months.
To address the concerns, Booz Allen said the Velocity edition lays out a systems-level approach to deploying agentic AI securely, spanning the full AI lifecycle rather than treating security as a one-time checklist. The report emphasizes identity governance, continuous monitoring, red teaming, formal validation, and the use of zero trust principles.
Identity governance refers to managing who or what is allowed to access systems and data, including permissions for AI-driven workflows. Continuous monitoring involves tracking performance and behavior after deployment, while red teaming and formal validation are designed to test an AI system against adversarial scenarios and to confirm it meets defined requirements. Zero trust principles shift security away from implicit trust inside networks and toward ongoing verification.
Booz Allen said the survey was conducted in partnership with Market Connections and involved an online survey of 105 federal government decision makers and influencers involved in IT and cybersecurity strategy. The fieldwork was conducted in April 2026.
For Booz Allen, the survey is positioned as both a warning and a planning tool for organizations wrestling with how to gain operational benefits from agentic AI without losing control of what those systems do and how they can be held accountable. The company described the underlying theme as a challenge for cybersecurity leaders: harnessing speed and autonomy while ensuring security and accountability.
Still, the release does not detail what specific agentic AI systems are being piloted, how agencies define “high confidence,” or what security gaps respondents are most concerned about. It also does not break out results by agency type, mission area, or procurement status, limiting how far the findings can be used to pinpoint where implementation risk is concentrated.
Looking ahead, the key question for federal adopters will be whether security measures such as identity governance and continuous monitoring keep pace with expanding deployments. As agencies push agentic tools from pilot into production, Booz Allen’s survey suggests the deciding factor may be governance maturity, including validated controls, testing discipline, and ongoing verification of AI behavior in operational environments.
Why It Matters
- The gap between deployment and security confidence suggests agencies may need faster progress on governance, testing, and operational controls to match adoption timelines.
- High concern about AI-accelerated cyber threats indicates pressure to treat agentic AI as both a defensive opportunity and a risk factor in threat models.
- If agencies proceed with autonomy, demand may rise for security frameworks that can demonstrate accountability and validated performance rather than relying on vendor promises.
Sources
Key Facts
- Booz Allen released a survey on agentic AI adoption in federal IT and cybersecurity environments.
- 58% of 105 federal IT and cybersecurity decision makers surveyed said their agencies have deployed or are piloting AI agents.
- 28% said they have high confidence in deploying AI agents securely.
- 79% of respondents reported being “very” or “extremely” concerned about AI-enabled adversaries accelerating cyberattacks in the next 12 to 18 months.
- The survey was fielded in April 2026 and conducted with Market Connections.
- Booz Allen said its accompanying Velocity edition highlights security controls across the agentic AI lifecycle, including identity governance, continuous monitoring, red teaming, formal validation, and zero trust.
Defense Related
U.S. Army deal ties up $54B in Patriot missiles, underscoring demand for stacked air defenses
A newly reported purchase of roughly $54 billion in Patriot air-defense missiles from Lockheed Martin points to continued urgency for both high- and lower-tier intercept capabilities, as threats in the Middle East keep pressuring U.S. procurement.
Archer Aviation executive says Boeing deal could “significantly change the profile” as Insitu’s $200 million revenue business joins
The comments, reported by Yahoo Finance, point to a potentially broader platform for the eVTOL company after a Boeing-linked transaction and the addition of Insitu’s reported $200 million revenue business line.
Boeing tie-in complicates Archer’s eVTOL path as Joby leans harder into commercial air-taxi buildout, market report says
A new market-focused comparison of Archer Aviation and Joby Aviation suggests the strategies are diverging, even as both pursue electric vertical takeoff and landing aircraft for passenger air taxis.
Jim Cramer calls Lockheed Martin “sensational” in Mad Money lightning-round exchange
During the August 6 episode of Mad Money, a caller asked about Lockheed Martin’s long-term outlook, prompting Jim Cramer to deliver a brief but upbeat endorsement.
Archer Aviation highlights growth, new Boeing-backed strategy, and AI push on Q2 earnings call
The electric aircraft developer reported sharp revenue growth in Q2 and used its earnings call to outline a larger partnership direction involving Boeing, alongside a newly described AI platform. The company also emphasized ongoing cash burn as it pursues certification and scaling.
Boeing BA ties up three aviation units in an equity-style deal with Archer, per market reports
Boeing is reportedly transferring Wisk Aero, SkyGrid and Insitu to Archer Aviation in an arrangement that would give Boeing a significant stake in Archer and a governance role, according to a Yahoo Finance report. The move adds clarity to how Boeing plans to monetize parts of its portfolio beyond its core aircraft business.
Boeing buys into Archer Aviation, taking a 20% stake as ACHR shares surge
A deal described in market coverage ties Boeing’s investment plans to Archer Aviation’s electric air-taxi efforts, sending Archer’s stock sharply higher. The move underscores how legacy aerospace firms are looking for footholds in the fast-changing eVTOL market.
Lockheed Martin and Albany Engineered Composites announce teaming deal aimed at scaling hypersonics production
The defense contractor is pairing its systems integration and battle-ready engineering capabilities with Albany’s agile composite manufacturing to pursue full-rate hypersonic production opportunities.
L3Harris completes hot-fire test milestone for Missile Defense Agency Next Generation Interceptor motor
The defense contractor said it successfully ran a hot-fire test of the second stage of an advanced large solid rocket motor tied to the U.S. Missile Defense Agency’s Next Generation Interceptor program.
Archer Aviation reports second-quarter 2026 results and says Boeing partnership targets “physical AI” for aerospace and defense
The eVTOL developer said its latest quarter advanced company operations while also announcing a deal with Boeing aimed at applying physical artificial intelligence concepts to aircraft and defense systems.