Activities
Divisions
Programs
Activities
Divisions
Programs
Study reveals that hackers can effortlessly decipher your conversations with ChatGPT and other AI platforms
Israeli researchers have discovered significant privacy flaws in numerous AI chatbots, such as ChatGPT. Even though developers are implementing encryption, hackers can still infiltrate using side-channel attacks.
A new research conducted by scientists at Ben-Gurion University in Israel has highlighted substantial privacy risks present in numerous AI chatbots, sparking worries over the confidentiality of personal discussions.
Yisroel Mirsky, who leads the Offensive AI Research Lab at Ben-Gurion University, suggests that these security flaws can be manipulated by ill-intentioned individuals to secretly listen in on conversations held on platforms such as ChatGPT.
Mirsky emphasized that anyone using the same Wi-Fi or local area network (LAN) as those involved in the chat, or even distant harmful individuals, can secretly observe and eavesdrop on conversations.
The study identifies these vulnerabilities as "side-channel attacks," a technique where third parties collect information indirectly through metadata or other non-direct ways instead of breaking through security defenses.
Rather than breaking through firewalls as standard hacks do, side-channel attacks exploit vulnerabilities in encryption systems. Even with encryption measures put in place by AI developers such as OpenAI, the team led by Mirsky found loopholes in their encryption execution, making the content of messages prone to being intercepted.
Although side-channel attacks are usually less intrusive, they present considerable threats, as shown by the scientists' capability to predict chat cues with a precision of 55 per cent. This vulnerability lets harmful individuals quickly identify sensitive subjects.
The research primarily examines the encryption methods used by OpenAI, implying that most chatbots, with the exception of Google's Gemini, are vulnerable to similar security breaches.
At the heart of these weaknesses is the deployment of "tokens" by chatbots, which aid in effective interaction between users and AI models. Even though chatbot messages are usually encoded, the tokens present a security risk that was not noticed before.
Having real-time access to token data allows harmful individuals to deduce conversation cues, similar to eavesdropping on a conversation behind a closed door.
In order to validate their results, the group led by Mirsky used another AI model to assess unprocessed data gathered via the side-channel. Their trials showed a high likelihood of accurately forecasting conversation prompts, highlighting the seriousness of the security flaw.
In response to these worries, Microsoft reassured users that the vulnerability impacting its Copilot AI would not likely lead to personal information being breached. Nevertheless, the firm committed to rapidly resolving the problem by providing updates to protect its clients.
These weaknesses bear significant consequences, especially when considering delicate matters like abortion and LGBTQ matters, where confidentiality is crucial. Taking advantage of these weaknesses could lead to severe repercussions, possibly putting at risk those looking for information on these subjects.
The growing discussions on artificial intelligence ethics and privacy highlight the critical necessity for strong security protocols to safeguard user privacy during interactions powered by AI.
Look for us on YouTube
Best Programs
Associated Narratives
can be found on YouTube
Firstpost holds all rights and protections under copyright law as of 2024
+ There are no comments
Add yours