NITDA Issues Cybersecurity Alert Over Vulnerabilities in OpenAI GPT-4.0, GPT-5 Models

NITDA Issues Cybersecurity Alert Over Vulnerabilities in OpenAI GPT-4.0, GPT-5 Models

The National Information Technology Development Agency (NITDA) has issued a fresh cybersecurity warning to Nigerians regarding critical vulnerabilities in OpenAI’s GPT-4.0 and GPT-5 series, which could expose sensitive data and compromise digital safety.

In a statement on Monday, NITDA’s Director of Corporate Affairs and External Relations, Mrs Hadiza Umar, confirmed the discovery of seven major weaknesses affecting the AI models. According to the agency, these vulnerabilities allow attackers to manipulate ChatGPT using a technique called indirect prompt injection.

NITDA explained that cybercriminals can hide malicious instructions within everyday online content, including webpages, social media comments, and shortened URLs. While users browse, summarise text, or use AI-powered search tools, ChatGPT may unintentionally execute these hidden commands.

READ ALSO:

• Parents, teachers slam Federal Ministry for forcing new subjects on SSS 3 students
• First Lady Clashes with Governor Adeleke as Videos Go Viral in Ile-Ife
• US Lawmaker Optimistic Nigeria Can Defeat Terrorism Amid Washington Security Talks

Some weaknesses enable attackers to bypass safety filters via trusted domains or exploit markdown rendering bugs to conceal harmful content. The agency also highlighted the risk of memory poisoning, where malicious prompts secretly alter ChatGPT’s behaviour across future interactions.

“Although OpenAI has addressed some of these issues, large language models still struggle to distinguish legitimate queries from harmful embedded instructions,” Umar said. She warned that the implications include unauthorised system actions, data leakage, manipulated AI responses, and long-term behavioural distortion.

NITDA urged Nigerians—particularly tech-driven businesses, digital creators, and young professionals—to adopt responsible AI usage practices. The agency stressed that as AI continues to shape sectors like entertainment, sports, and technology, prioritising digital safety is crucial to prevent exploitation by cybercriminals.

NITDA Issues Cybersecurity Alert Over Vulnerabilities in OpenAI GPT-4.0, GPT-5 Models