In today's rapidly evolving digital landscape, the ability to create and distribute your own GPT has become astonishingly accessible. Thanks to OpenAI's Custom GPTs and the GPT store, artificial intelligence tooling is more democratized than ever.

However, the ease of crafting prompts for these GPTs is a double-edged sword. While you can tailor the AI to specific needs, other users may potentially access the inputs, files, or instructions provided to the Custom GPT.

Pro's

One of the remarkable features of GPTs is their ability to consume large amounts of information. By uploading files filled with data, users can enrich the AI with knowledge and instructions, turning it into a specialized tool.

This allows a more customized interaction, including asking questions about content that is not inherently available in the base model.

The Custom GPT prompt leaking

There is a significant risk of user data access. If a user of your custom GPT can access sensitive data provided to the system, this raises serious data privacy and security concerns.

In some situations, uncovering the underlying mechanics of a custom GPT model can be as straightforward as asking for the instructions provided by the creator. With the right phrasing, an AI may reveal foundational guidelines it operates on.

Risk assessment

This prompt injection is labeled as RISK: HIGH IMPACT: HIGH . If creators only use public data, the risk remains more contained. If sensitive or proprietary information is uploaded, the consequences can be severe.

This scenario highlights the importance of strict data governance and awareness within organizations using custom GPT models.