Security Vulnerability Discovered in iOS AI Apps

Researchers have found in a recent study that 282 out of 444 tested AI chatbot apps for iPhone expose API keys and access to OpenAI proxies through their network traffic. This corresponds to nearly two-thirds of the applications examined. The security vulnerability allows attackers to access paid AI services simply by monitoring the apps' network traffic. The analysis revealed that in many cases, access to the AI services was visible through the transmission of unencrypted API keys, reusable tokens, or backend servers that accept requests without authentication. These vulnerabilities could be exploited by malicious actors to send requests to the AI models on behalf of the developers.

The researchers tested the apps to assess the security standards in the field of artificial intelligence. The results show that many developers do not pay sufficient attention to the security of their applications. The study highlights that the use of unencrypted connections and inadequate implementation of authentication mechanisms can lead to significant security risks. Another concerning aspect is that most of the affected apps do not provide warnings or indications of the security vulnerabilities. Users could unknowingly jeopardize their data and account information by using these apps.

Researchers recommend that developers urgently take measures to improve security. The study points out that the vulnerabilities affect not only the developers but also the users of the apps. A successful attack could lead to financial losses and a loss of trust in the affected applications. The researchers call for a comprehensive review of security practices in app development.

To enhance the security of the apps, the researchers recommend implementing encrypted connections and using strong authentication mechanisms. Developers should ensure that API keys and tokens are not transmitted in plaintext. Additionally, regular security audits and penetration tests should be conducted to identify potential vulnerabilities early. The study also showed that awareness of security issues among developers and users needs to be improved. Training and information campaigns could help raise awareness of the importance of IT security.

The researchers emphasize that a proactive approach to security issues is crucial to prevent future incidents. The results of the study were presented at a cybersecurity conference held on June 30, 2026. Industry experts discussed the challenges and solutions in the field of app security. The researchers hope that their findings will lead to a broader discussion about security standards in software development.

The security vulnerability affects not only the developers but also the users who rely on the integrity of the applications. The researchers urge the affected companies to take immediate action to ensure the security of their products. According to the study's findings, up to 50,000 users could be affected by the security vulnerabilities. The study is expected to be published in a scientific journal to make the results accessible to a wider audience. The researchers hope that their work will contribute to raising security standards in app development and strengthening user trust in AI applications.