Immersive leadership training, and training on safe AI use, OPSEC and cybersecurity

Last updated October 21, 2026

Simulations in which leaders practice difficult conversations, and hands-on training in using AI tools and handling sensitive information safely.

This guide lists 11 providers of immersive leadership, safe AI use and cybersecurity training, in alphabetical order: Bodyswaps, Cornerstone (Talespin), Hack The Box, Lockheed Martin, Moth+Flame, Mursion, SimSpace, Skillsoft, Strivr, USC Institute for Creative Technologies and VirtualSpeech. The order is not a ranking. Each provider is described with the same fields, from its own materials and public records, with sources.

1. Questions buyers ask

2. Overview

This category covers two kinds of immersive training. One lets officers, NCOs and civilian supervisors practice leadership conversations, such as counseling and feedback, with virtual humans or AI role-players. The other puts personnel in simulated networks, cyber ranges and hands-on labs to practice cybersecurity and the safe use of AI tools.

Army doctrine defines leader development as a deliberate, continuous and progressive process and says a skill is learned through repeated, deliberate practice. Simulated conversations give leaders a low-stakes way to rehearse hard talks before having them with real people.

Generative AI is now part of daily work. On Dec. 9, 2025, the Secretary of War launched GenAI.mil for military personnel, civilians and contractors and directed every member of the Department to learn it. The Navy's rules for GenAI.mil bar protected health information and personally identifiable information.

The problem

Army doctrine calls counseling one of a leader's most important development responsibilities and says skills improve through repeated, deliberate practice. Federal rules require agencies to train new civilian supervisors within one year of appointment and at least every three years after.

Navy guidance warned in 2023 that large language models save every prompt, creating a persistent OPSEC risk, and that combining prompts and outputs could release sensitive or classified information. It also warned that generative AI can produce fabricated 'hallucinations' that need human review.

Required cyber awareness training is an online course taken at in-processing and every year, and OPSEC awareness training is also annual. Hands-on skills need practice environments: cyber ranges let cyber teams rehearse attack and defense where mistakes carry no real-world consequences.

Mandates and programs

• Army leader development doctrine and federal supervisor training (FM 6-22; AR 350-1; 5 CFR 412.202). FM 6-22 (2022, updated 2025) sets Army leader development doctrine, including developmental counseling, and AR 350-1 (June 2025) governs training and leader development for Soldiers and Army Civilians. 5 CFR 412.202 requires training for new federal supervisors within one year and every three years.
• Annual cyber awareness, OPSEC and CUI training (AR 350-1; AR 530-1 implementing DoDD 5205.02E). AR 350-1 lists cyber awareness training at in-processing and annually, and annual OPSEC and CUI refresher training, as mandatory. AR 530-1 implements DoD Directive 5205.02E and requires OPSEC awareness training within 30 days of arrival and every year.
• GenAI.mil and Department of War direction to adopt generative AI (December 2025). The Secretary of War's Dec. 9, 2025 memo launched GenAI.mil and directed all personnel to learn it. The Navy then designated GenAI.mil its mandated CUI and IL5 generative AI platform, barred PHI and PII on it, and listed free AI training resources.
• OMB memos M-25-21 and M-25-22 (April 3, 2025). M-25-21 says the federal workforce has a responsibility to maintain at least foundational knowledge of using AI responsibly and requires periodic training for operators of high-impact AI; M-25-22 governs agencies' AI acquisitions. Neither applies to AI used as a component of a national security system.
• AI Training Act (Public Law 117-207, 2022). The AI Training Act requires OMB to provide an AI training program on AI capabilities and risks for executive-agency program, engineering, contracting, logistics and cost staff, updated at least every two years. Its definition of executive agency excludes the Department of Defense.

3. Providers compared

Every provider below is described with the same fields and the same length limits, from its own published materials and public records. Where a provider does not publish something, the entry says “Not published.” Providers are in alphabetical order.

4. What to ask a vendor

1. What does the learner actually do: speak, type, or act in VR, and how many repetitions can they get? Army doctrine says learning a skill takes repeated, deliberate practice: an attempt, analysis of the result, and another attempt using the improvements.
2. Are scenarios built on our doctrine, such as FM 6-22 counseling or our unit's OPSEC critical information? Scenarios that match your doctrine and mission let leaders and staff practice the conversations and decisions they will actually face.
3. If AI plays the role-player, which model runs it, where are conversations stored, and is our data used for training? Navy guidance warned that large language models save prompts, creating a persistent OPSEC risk, so buyers need to know where practice data goes.
4. How are AI responses checked for accuracy and fabricated content? Navy guidance notes generative AI tools can produce hallucinations and says they must be paired with human review.
5. Where can it run: browser, mobile, offline or air-gapped, and under which security authorization or impact level? Cloud services used by agencies are authorized through programs such as FedRAMP, and the Navy has mandated GenAI.mil, authorized up to IL5 for CUI, as its generative AI platform.
6. What is scored, and can scores and after-action data be exported to our learning system? Ask for behavior-level or operational scores, not only completion rates, so you can see whether practice is changing performance.
7. Which contract vehicle can we use, such as a GSA schedule or NASA SEWP? GSA's Multiple Award Schedule gives government buyers commercial products and services at pre-negotiated prices, and NASA SEWP is a government-wide contract for IT products and services.

5. How buying works

Many products in this category are commercial software subscriptions. GSA's Multiple Award Schedule offers federal, state, local and tribal buyers commercial products and services at pre-negotiated prices, and NASA's SEWP government-wide acquisition contract provides IT products and services to all federal agencies.

Public contract records show units buying both directly and through resellers: Air Force contracting offices have bought avatar-based simulations directly, and Air Force units have renewed cyber range licenses through resellers' SEWP contracts.

Large cyber training environments are bought as programs. The Army's Persistent Cyber Training Environment is built and sustained under Army contracts to train the Cyber Mission Force for U.S. Cyber Command, and the National Cyber Range runs under Army contracts for the Test Resource Management Center.

Cloud tools that hold government data need security authorization. FedRAMP certifies cloud services for federal use and lists them in a public marketplace; for generative AI, the Navy directs users to GenAI.mil, which is authorized for CUI up to IL5.

For SBIR, GSA schedules, OTAs and other routes in general, see How buying works.

6. More questions