National Information Technology Development Agency cautions public servants against ChatGPT and Gemini security risks
The National Information Technology Development Agency (NITDA) issued a strict cybersecurity advisory in Abuja on Saturday, warning all organisations and their employees against inputting classified, personal, or confidential information into public artificial intelligence systems.
The alert, disseminated via the agency emergency response team, highlights severe data protection and security vulnerabilities linked to the adoption of popular public Large Language Models such as ChatGPT, Gemini, Claude, and Copilot.
By utilizing these third-party platforms for routine drafting and research, personnel risk exposing sensitive details, according to information published by The Sun Nigeria.
Data retention risks in public Large Language Models
NITDA explained that any information submitted to public artificial intelligence platforms might permanently escape the direct control of the originating organization. Service providers frequently retain, log, or utilize this transmitted data to train their underlying machine learning models.
When employees leverage these systems for daily office duties, they can inadvertently disclose Personally Identifiable Information (PII) and internal records. This unauthorized exposure directly violates established data privacy mandates and legal frameworks.
Legal and national security implications of AI breaches
The regulatory authority emphasized that exposing official government documents or confidential corporate details to external providers compromises national security and public trust. Such security lapses can trigger severe disciplinary procedures and legal penalties for both workers and their employers.
Furthermore, mishandling personal records through digital platforms constitutes a formal data breach, exposing institutions to statutory fines under current legislation.
Recommended mitigation strategies for workplace AI usage
To safely navigate emerging digital threats, NITDA outlined several mandatory precautions for corporate entities and government institutions across the country:
- Avoid entering confidential, classified, or personal data into public artificial intelligence interfaces.
- Utilize exclusively approved internal tools for official duties.
- Thoroughly anonymize, pseudonymize, or redact sensitive information before interacting with external platforms.
- Examine platform privacy policies and avoid uploading internal files without formal authorization.
Broader cybersecurity concerns and regulatory updates
This recent advisory arrives amid mounting anxieties regarding workplace technology risks, following previous alerts concerning malicious software capable of harvesting browser credentials. Concurrently, the Nigeria Data Protection Commission (NDPC) prepares to update national frameworks to properly govern artificial intelligence, big data, and robotics.
Frequently Asked Questions
Why did NITDA issue a warning about public AI tools?
The National Information Technology Development Agency warned workers because entering confidential, classified, or personal data into platforms like ChatGPT and Gemini poses serious data privacy and national security risks.
What specific public AI models are affected by the NITDA advisory?
The advisory specifically mentions popular public Large Language Models including ChatGPT, Gemini, Claude, and Microsoft Copilot.
How can employees protect sensitive data while using artificial intelligence?
Employees should use only approved corporate tools, avoid uploading internal documents without permission, and thoroughly anonymize or pseudonymize any sensitive information before interaction.
What are the consequences of exposing data on public AI platforms?
Exposing sensitive data can lead to personal data breaches, legal penalties, national security compromises, and severe disciplinary actions for both employees and their organizations.
