Human-machine teaming in DoD and NATO guidance

Official DoD, NATO, UK MOD and EASA pages talk about human judgment, human responsibility, human control and, in one EASA notice, human-AI teaming. They do not, on the sources below, publish a single numbered “human-machine teaming requirements” checklist that applies to every AI system. This article collects the sentences those pages actually use.

Status as of 14 October 2026: NATO’s 22 October 2021 strategy summary and revised strategy of 10 July 2024 are still live; the 2024 page still restates the six Principles of Responsible Use, including Responsibility and Accountability and Governability. GOV.UK still dates JSP 936 Part 1 Published 13 November 2024. EASA’s 6 March 2024 news release still describes Level 2 as human-AI teaming. Live automated fetches of war.gov and ai.mil returned HTTP 403 this pass; DoD ethical-principles and CDAO toolkit sentences are kept from those same URLs.

United States — Responsible, Governable, and 3000.09

The Department of Defense adopted five AI ethical principles in February 2020. Two of them are about people, not models:

• Responsible. DoD personnel will exercise appropriate levels of judgment and care while remaining responsible for the development, deployment, and use of AI capabilities.
• Governable. The Department will design AI so that it has the ability to detect and avoid unintended consequences, and the ability to disengage or deactivate deployed systems that demonstrate unintended behavior.

Those sentences apply to AI capabilities, not only to weapons. For weapons, DoD Directive 3000.09 adds a narrower rule: systems will be designed to allow commanders and operators to exercise appropriate levels of human judgment over the use of force. The same paragraph requires human-machine interfaces that are readily understandable to trained operators, that provide transparent feedback on system status, and that provide clear procedures for trained operators to activate and deactivate system functions. Deputy Secretary Hicks’s public line is There is always a human responsible for the use of force, full stop.

3000.09 does not apply to systems that are not weapon systems. See Autonomy and DoD Directive 3000.09 explained.

The CDAO toolkit, released 14 November 2023, is a voluntary process that identifies, tracks, and improves alignment of AI projects to RAI best practices and the DoD AI Ethical Principles. The Department’s release says it offers guidance and an RAI standard to current and future DoD industry partners. We have not found a source that makes completing the toolkit a contractual human-machine-teaming clause. See DoD Responsible AI.

NATO — Responsibility and Accountability, and Governability

NATO’s 22 October 2021 strategy summary places the Principles of Responsible Use for AI in Defence at the front of the strategy. The principles apply across all types of AI applications and do not affect or supersede existing obligations and commitments, both national and international. The revised strategy of 10 July 2024 restates the same six names.

Two principles name humans:

• Responsibility and Accountability. Applications will be developed and used with appropriate levels of judgment and care; clear human responsibility shall apply in order to ensure accountability.
• Governability. They will be developed and used according to intended functions, allowing appropriate human-machine interaction, the ability to detect and avoid unintended consequences, and disengagement or deactivation … when such systems demonstrate unintended behaviour.

That is the official NATO wording for human-machine interaction. The 2021 summary does not, on the page we cite, define a staffing ratio, a user-interface standard, or a certification test for a “team.” NATO also says some Allies already have national principles of responsible use. See NATO PRUAI.

United Kingdom — Human-centricity and human control

UK MOD JSP 936 Part 1 (published 13 November 2024) mandates the application of ambitious, safe and responsible practices relating to all Defence projects that include Artificial Intelligence (AI). Where the JSP says must, the action is a compulsory requirement for the MOD audience it names.

The JSP copies five ethical principles verbatim from the Ambitious, Safe and Responsible policy. Two of them are about people:

• Human-centricity. The impact of AI-enabled systems on humans must be assessed and considered, for a full range of effects both positive and negative across the entire system lifecycle.
• Responsibility. Human responsibility for AI-enabled systems must be clearly established, ensuring accountability for their outcomes, with clearly defined means by which human control is exercised throughout their lifecycles.

ASR says the principles do not affect or supersede existing legal obligations. JSP 936 does not provide a list of AI-enabled systems or capabilities which are compliant or not compliant. See UK MOD JSP 936.

EASA — Level 2 as human-AI teaming

EASA’s 6 March 2024 news release for Concept Paper Issue 2 says Issue 2 refines guidance for Level 1 AI applications, those enhancing human capabilities, and provides guidance for Level 2 systems built on human-AI teaming. That is the official EASA use of the teaming phrase on the pages we cite. The paper itself does not constitute final detailed guidance and covers Level 1 and Level 2 AI applications but not yet Level 3.

EASA’s wording is aviation guidance, not a DoD or NATO contract clause.

What we searched for and did not find

We looked, on the official pages in the source list, for a single DoD or NATO document titled as human-machine teaming requirements with testable interface, staffing or training clauses that apply to all AI. We did not find one. What those pages require is:

Questions

Is there a DoD “human-machine teaming” standard that contractors must certify to?

Not on the official pages we cite. DoD’s public human-facing requirements for AI are the ethical principles and, for weapon systems, DoDD 3000.09. The CDAO toolkit is voluntary.

Does NATO require a human in every AI decision?

NATO requires clear human responsibility and appropriate human-machine interaction. The 2021 summary does not define that as a human approving every output. The principles do not supersede national law.