How to secure the use of artificial intelligence in your company?
The dangers of information leaks linked to the use of AI-generated text
AI-generated text has revolutionised the way in which users interact with computers and digital devices, enabling them to carry out complex tasks more rapidly and more efficiently than ever before. However, its use can also present risks for the security of data and of industrial secrets.
The risks for data security
Involuntary disclosure of sensitive information:
Employees can accidentally disclose sensitive or confidential information to the AI text generator. The data may then be stored in databases accessible to third parties.
Phishing attacks:
AI texts can be used for sophisticated attacks via phishing, where apparently legitimate messages are sent to users encouraging them to disclose sensitive information or click on malicious links.
Intentional data leaks:
Ill-intentioned users can use AI texts to extract confidential information and disclose it to third parties.
Security vulnerabilities:
AI servers may be vulnerable to attacks, enabling malicious parties to access sensitive information stored in the systems (conversation saved between the user and the artificial intelligence).
DATIVE advice for mitigating the risks:
- Training and awareness-raising for employees
- Strict security policy with limitation of usage
- Sandboxing of development networks
- Sign up for a vulnerabilities monitoring service