Academic Branch — Nigerian Defence Academy

Nigerian Defence Academy: Artificial Intelligence Policy

Directorate of Academic Planning

October 23, 2026

Nigerian Defence Academy: Artificial Intelligence Policy

Nigerian Defence Academy: Artificial Intelligence Policy
Draft for consideration by the Academic Board and Academy Management. 

Policy No.: [NDA/AI/001] | Version 0.1 

1. Purpose
This policy sets out how the Nigerian Defence Academy (NDA) will adopt, govern and use artificial intelligence (AI) in teaching, research, administration and officer cadet development. It aims to:
Enable responsible innovation that strengthens academic quality, operational efficiency and national security capability.
Protect classified, personal and institutional information.
Preserve academic integrity, discipline and the professional ethics expected of military officers.
Ensure that people, not machines, remain accountable for decisions.
2. Scope
This policy applies to all NDA staff (military and civilian), academic and administrative personnel, officer cadets, postgraduate students, visiting lecturers, researchers, contractors and partners who use or develop AI systems on behalf of, or using the resources of, the Academy.
"AI system" means any software or service that generates content, predictions, recommendations, classifications or decisions from data, including generative AI tools (such as chatbots and image generators), machine learning models, computer vision systems, automated decision-support tools and AI features embedded in other software.
The policy covers AI used on Academy networks and devices, and personal tools when used for Academy work.
3. Guiding Principles
Human accountability. A named person remains responsible for every AI-assisted output, decision and recommendation. AI cannot be blamed or credited in place of a person.
Security first. Information is handled according to its classification. Nothing sensitive is entered into AI tools that the Academy has not approved for that level of information.
Integrity and honesty. AI use in coursework, research and official work must be disclosed where this policy requires it, and must never be used to deceive.
Fairness and non-discrimination. AI must not be used in ways that disadvantage individuals or groups on the basis of ethnicity, religion, gender, region, disability or social background.
Transparency and explainability. Where AI materially influences a decision about a person, the basis for the decision must be explainable and open to review.
Proportionality. The level of control applied to an AI use is proportionate to the risk it carries (see Appendix A).
Capability and literacy. The Academy will build AI skills across its community, so that use is informed rather than prohibited by default or adopted uncritically.
Compliance. AI use must comply with Nigerian law, including the Nigeria Data Protection Act 2023, the Cybercrimes (Prohibition, Prevention, etc.) Act, applicable Armed Forces regulations and the Official Secrets Act, together with any national AI policy or strategy in force.
4. Governance and Responsibilities
4.1 AI Governance Committee. The Academy will establish an AI Governance Committee, chaired by the Academy Provost and comprising representatives of the Faculties (including Computer Science), the Directorate of ICT, Legal Services, the Registry, Library, Security and Intelligence, Student Affairs and the Cadet Wing. The Committee will:
approve and maintain the register of approved AI tools and use cases;
review high-risk proposals and decide on exceptions;
oversee training, monitoring and incident response;
report to the Academic Board and Academy Management at least twice a year;
review this policy annually.
4.2 Heads of Department and Units. Ensure staff in their units understand and follow this policy, approve low-risk AI use within their areas, and refer medium- and high-risk proposals to the Committee.
4.3 Directorate of ICT / Department of Computer Science. Evaluates tools for security and interoperability, maintains the approved tools list, configures access controls, and monitors for unauthorised use.
4.4 Data Protection Officer. Advises on data protection impact assessments, handles data subject requests relating to AI processing, and liaises with the Nigeria Data Protection Commission where required.
4.5 Individual users. Use AI only as permitted by this policy, verify outputs, protect information, disclose use where required, and report incidents promptly.
5. Teaching and Learning
5.1 Staff use. Lecturers may use approved AI tools to prepare teaching materials, generate examples, draft feedback and design activities, provided that they verify accuracy, remove sensitive or personal data from inputs, and remain responsible for the content delivered.
5.2 Course-level AI statements. Every course must state at the outset which of the following levels applies to each assessment:
Level 0, No AI: AI tools are not permitted.
Level 1, Limited AI: AI may be used for defined supporting tasks (for example brainstorming, language checking), with disclosure.
Level 2, Open AI with disclosure: AI may be used extensively, with a declaration of how it was used and critical evaluation of its output.
Where no statement is made, Level 0 applies by default to assessed work.
5.3 Student and cadet use. Cadets and students are encouraged to develop AI literacy. They must follow the course-level statement, must not submit AI-generated work as their own where it is not permitted, and remain fully responsible for the accuracy of anything they submit.
5.4 Assessment design. Departments should favour assessment methods that demonstrate real understanding, such as supervised practicals, oral examinations, in-class tasks, project work with viva voce and process-based evidence, rather than relying on AI-detection software.
6. Academic Integrity
6.1 The following are breaches of academic integrity: submitting AI-generated material as one's own work where not permitted; failing to disclose AI use where disclosure is required; fabricating references or data produced by AI; and using AI to impersonate another person or to circumvent examination controls.
6.2 AI-detection tools are unreliable and must not be the sole basis for an allegation. Suspected misconduct must be investigated under the Academy's existing disciplinary procedures, taking account of drafts, version history, interviews and assessment of the student's understanding.
6.3 Penalties for AI-related misconduct follow the same scale as other forms of academic misconduct. For officer cadets, such conduct is also a matter of honour and discipline and may be dealt with under Cadet Wing regulations.
7. Research
7.1 Researchers may use AI for literature exploration, coding assistance, data analysis and drafting, subject to this policy and the requirements of funders, journals and ethics committees.
7.2 AI must not be listed as an author. Researchers must disclose substantive AI use in methods or acknowledgements, verify all citations, and take responsibility for the integrity of their results.
7.3 Research involving human participants, biometric data, facial or image datasets, or surveillance-related AI (including object detection and computer vision applied to people or sensitive sites) requires ethics approval and a data protection impact assessment before data collection begins.
7.4 Unpublished research, grant proposals, peer-review material and supervised student work must not be uploaded to public AI tools.
7.5 Datasets must be documented, lawfully obtained, assessed for bias and stored securely. Research with dual-use or defence implications must be reviewed by the Committee before publication or external collaboration.
8. Administration and Operations
8.1 AI may support administrative functions such as drafting correspondence, summarising documents, scheduling, timetabling, helpdesk services and data analysis.
8.2 AI must not make final decisions on admissions, recruitment, promotion, discipline, grading, welfare, security clearance or any other matter materially affecting a person's rights or career. AI may inform such decisions only where a qualified person reviews the evidence, can override the output, and records their reasoning.
8.3 Result processing, transcripts and student records systems must not incorporate AI that alters or generates academic records without Committee approval and a documented audit trail.
9. Information Security, Classification and Data Protection
9.1 Classified information. Information classified at any level, and information that is operationally, institutionally or personally sensitive, must never be entered into public or unapproved AI tools. Only tools approved by the Committee for the relevant classification may process such information, and only within Academy-controlled environments.
9.2 Personal data. Processing of personal data by AI systems must be lawful, fair, limited to a specific purpose and minimised. Staff and students will be informed when their data is processed by AI in ways that affect them. Where processing is likely to pose a high risk, a data protection impact assessment is required.
9.3 Public AI tools. When using public generative AI services, users must not input names, service numbers, biometric data, health data, examination material, financial data, personnel records, security plans or any non-public Academy information.
9.4 Approved tools. Only tools on the Academy's approved register may be used for official work above the lowest risk tier. The register will record each tool's purpose, owner, data permitted, hosting location and review date.
9.5 Accounts and devices. Official AI use should be through Academy-managed accounts where available. Users must not install unapproved AI applications, browser extensions or plug-ins on Academy systems.
9.6 Retention. Prompts, outputs and logs relating to official decisions or records are institutional records and must be kept according to Academy retention schedules.
10. Defence, Security and Military Applications
10.1 Any AI system developed, procured or tested for operational, intelligence, surveillance, targeting, weapons-related or command-and-control purposes is automatically classified high risk and requires approval of Academy Management and, where applicable, the relevant Service or Defence authorities.
10.2 The Academy affirms that meaningful human control must be preserved over any use of force, and that AI will not be used to make autonomous decisions about the use of lethal force.
10.3 Officer cadets will be educated on the legal, ethical and operational implications of military AI, including the laws of armed conflict, risks of automation bias, adversarial attacks, disinformation and deepfakes.
10.4 Collaboration with foreign institutions, companies or governments on AI involving defence-relevant data, models or technology requires prior security and legal clearance.
11. Fairness, Bias and Human Oversight
11.1 Before an AI system is used in a way that affects people, its developers or sponsors must assess the risk of bias, including performance differences across languages, accents, skin tones, genders and regions, and document the mitigation measures taken.
11.2 Users must treat AI outputs as unverified. Outputs must be checked for accuracy, bias, fabricated information and copyright or licensing issues before use.
11.3 Individuals affected by an AI-assisted decision have the right to request a human review and an explanation.
12. Intellectual Property
12.1 Users must respect copyright and licence terms when using AI tools and training data.
12.2 Ownership of AI-assisted work produced in the course of employment or study follows the Academy's existing intellectual property policy. Users must check tool terms to ensure that the Academy's rights and confidentiality are not compromised.
12.3 Proprietary Academy material, teaching content and research data must not be used to train external AI models without written approval.
13. Procurement and Development
13.1 All procurement of AI products or services must include a security review, data protection assessment, vendor due diligence (including data hosting location, sub-processors and incident history) and an exit plan.
13.2 AI systems developed in-house or by Academy researchers for deployment must be documented (purpose, data sources, testing and known limitations), tested before release and monitored after deployment.
13.3 Preference will be given, where feasible, to solutions that allow Academy control of data, local hosting and open standards, and that support Nigerian languages and contexts.
14. Training and AI Literacy
14.1 The Academy will provide AI awareness training for all staff and a core AI literacy component for all cadets and students, covering effective use, limitations, verification, privacy, security, ethics and academic integrity.
14.2 Advanced training will be provided for those who deploy, procure or oversee AI systems, including members of the Governance Committee, examiners and supervisors.
14.3 The Department of Computer Science will support the Committee with technical guidance, tool evaluation and curriculum development.
15. Monitoring, Incidents and Enforcement
15.1 AI use on Academy systems may be monitored in accordance with the Academy's ICT acceptable use policy and applicable law.
15.2 An AI incident includes any data leak or unauthorised disclosure through an AI tool, harmful or discriminatory output affecting a person, serious system failure, misuse of AI for fraud, harassment or disinformation, or any breach of this policy. Incidents must be reported to the Directorate of ICT and the Data Protection Officer without delay, and in any case within 24 hours of discovery.
15.3 Breaches may lead to disciplinary action under Armed Forces regulations, the Academy's staff and student conduct rules, or contract terms, and where appropriate referral to law enforcement.
15.4 Good-faith reporting of incidents and mistakes will be encouraged and will not itself attract sanction.
16. Review and Interpretation
This policy will be reviewed at least annually, and sooner if there are significant changes in technology, law, national policy or security circumstances. The AI Governance Committee is responsible for interpreting the policy. Questions should be directed to [contact office / email].
Appendix A: Risk Tiers
Tier
Examples
Requirements

Low
Grammar checking, brainstorming, summarising public material, drafting non-sensitive text
Approved or public tools permitted; no sensitive data; verify outputs

Medium
Teaching content generation, internal analytics, research data analysis, student support chatbots
Head of Department approval; approved tools only; data protection review; documented oversight

High
Admissions or grading support, recruitment, discipline, biometric or facial recognition, surveillance, anything involving classified data
AI Governance Committee approval; data protection impact assessment; security review; human decision-maker; audit logging

Prohibited
Autonomous use of lethal force; covert mass surveillance of staff or cadets; AI-generated impersonation or disinformation; entering classified data into unapproved tools; AI as sole decision-maker on a person's career or discipline
Not permitted under any circumstances

Appendix B: AI Use Disclosure Template
For coursework, research outputs and official documents where disclosure is required.
Tool(s) used and version:
Date(s) of use:
Purpose (for example, idea generation, editing, coding, data analysis):
Prompts or description of how the tool was used:
What I changed, verified or rejected:
Declaration: I confirm that I have reviewed all AI-assisted content and take full responsibility for it.
Name / Service number: ________  Signature: ________  Date: ________
Appendix C: Implementation Plan (Suggested)
Timeline
Action

Within 1 month
Approve policy; constitute AI Governance Committee; issue interim guidance to staff and cadets

Within 3 months
Publish approved tools register; require course AI statements; begin staff awareness training

Within 6 months
Launch cadet AI literacy module; complete data protection impact assessments for existing high-risk uses; set up incident reporting channel

Within 12 months
First annual review; report to Academic Board; update policy and training