Image: MLYearning
Northwestern University recently unveiled a critical vulnerability in custom Generative Pre-trained Transformers (GPTs).
Despite their versatility and adaptability, these advanced AI chatbots are prone to prompt injection attacks, risking exposure of sensitive information.
Custom GPTs, developed using OpenAI's ChatGPT and its GPT-4 Turbo Large Language Model, incorporate unique elements like specific prompts, datasets, and processing instructions for specialised tasks.
Image: Decrypt
However, these customisations and any confidential data used in their creation can be easily accessed by unauthorised parties.
An experiment by Decrypt demonstrated the ease of extracting a custom GPT's full prompt and confidential data through basic prompt hacking.
Testing over 200 custom GPTs, researchers found a high likelihood of such breaches, including the potential extraction of initial prompts and access to private files.
The study highlights two major risks: compromised intellectual property and breached user privacy.
Attackers can exploit the GPTs to either extract the core configuration and prompt ("system prompt extraction") or leak confidential training datasets ("file leakage").
Existing defences like defensive prompts prove ineffective against more sophisticated adversarial prompts.
The researchers argue for a more comprehensive approach to safeguard these AI models, emphasising that determined attackers can likely exploit current vulnerabilities.
The study calls on the AI community to develop stronger security measures, suggesting that simple defensive prompts are inadequate against such advanced exploitation techniques.
With the rising customisation of GPTs offering vast potential, this research serves as a critical reminder of the security risks involved.
Users are advised to exercise caution, especially with sensitive data, underscoring the need for enhanced AI security without compromising user privacy and safety.
The complete Northwestern University study is available for reading here.
OpenAI ramps up efforts with a successor to GPT-4, focusing on safety. Amid ethical scrutiny and criticism, a safety and security committee is formed to address concerns and ensure responsible AI development.
Huang BoOpenAI has made its ChatGPT accessible without the need for registration, aiming to democratize access to AI. While enhancing accessibility, the company ensures user safety through additional content safeguards. Despite facing challenges like slowing growth and legal disputes, this move reflects a broader trend of democratizing transformative technologies. However, uncertainties remain regarding AI's direction and potential misuse, raising concerns about deviation from original missions.
SanyaSince OpenAI launched the GPT Store in November last year, millions of GPTs have appeared for various scenarios such as office, education, life, and entertainment. GPTs are becoming assistants for many people to handle work, optimize writing skills, and obtain creative inspiration.
JinseFinanceThe OpenAI GPT Store faces pre-launch hurdles with unauthorized app replication, raising concerns on intellectual property protection and platform governance, posing uncertainties for developers awaiting guidance.
Xu LinIn the realm of evolving AI technology, WeHead boldly introduces the WeHead GPT Edition, priced at an astounding $5,000. This AI assistant stands apart by providing a physical face for users to engage with, sparking intriguing discussions about the future of AI companionship and the practicality of humanizing technology.
Xu LinJinseFinanceJinseFinanceJinseFinanceThe latest release transforms AI interactions by expanding the knowledge base to April 2023 and introducing support for 300-page documents.
JixuJinseFinance