Activities
Segments
Performances
Activities
Segments
Performances
Study reveals hackers' ability to intercept communications with ChatGPT, other AI services
A group of researchers from Israel discovered significant privacy weaknesses in various AI chatbots, including ChatGPT. Despite developers securing their services with encryption, hackers can still gain access through side-channel attacks.
A new research conducted by scientists at Ben-Gurion University in Israel has highlighted substantial privacy risks associated with various AI chatbots, leading to worries about the safety of confidential discussions.
Yisroel Mirsky, who leads the Offensive AI Research Lab at Ben-Gurion University, suggests that bad actors could take advantage of these weaknesses to spy on conversations carried out on platforms such as ChatGPT.
Mirsky pointed out that people using the same Wi-Fi or local area network (LAN) as those involved in a chat, or even distant harmful individuals, can secretly listen in and observe conversations without being noticed.
The study highlights these vulnerabilities as "side-channel attacks," a technique where third parties collect information indirectly through metadata or other non-direct sources instead of breaking through security defenses.
Rather than breaking through firewalls as conventional hacks do, side-channel attacks exploit vulnerabilities in encryption systems. Even with encryption measures put in place by AI creators such as OpenAI, Mirsky's group found defects in the way they implemented encryption, which made the content of messages prone to being intercepted.
Though side-channel attacks are typically less intrusive, they present serious dangers, as shown by the scientists' capacity to deduce chat cues with a 55 percent precision rate. This vulnerability renders sensitive subjects readily identifiable to harmful entities.
The research mainly examines the encryption methods of OpenAI, but it implies that apart from Google's Gemini, most chatbots are likely to be vulnerable to similar manipulations.
The crux of these weaknesses revolves around the employment of "tokens" by chatbots, which endorse smooth communication between users and AI systems. Even though chatbot communications are usually safeguarded through encryption, these tokens introduce a potential risk that was earlier ignored.
Gaining access to live token information allows harmful individuals to deduce dialogue cues, similar to eavesdropping a discussion behind a shut door.
In order to validate their results, the group led by Mirsky used another AI model to scrutinize raw information obtained via the side-channel. Their trials demonstrated a significant accuracy in forecasting dialogue prompts, highlighting the seriousness of the susceptibility.
Microsoft reacted to these worries by reassuring users that their personal information is not likely to be endangered by the vulnerability in its Copilot AI. Nevertheless, the company committed to resolving the problem swiftly with updates to protect its consumers.
The impact of these weaknesses is significant, especially on delicate matters like abortion and LGBTQ subjects, where confidentiality is crucial. Manipulating these weaknesses could lead to severe outcomes, possibly putting at risk individuals who are searching for information on these topics.
The discussions regarding the ethical aspects and privacy concerns of AI are becoming more heated. This highlights the immediate necessity for strong security protocols to guard users' privacy during interactions powered by AI.
Search for us on YouTube
Prime Programs
Associated Narratives
can be found on YouTube
Firstpost holds all rights, protected by copyright, as of 2024
+ There are no comments
Add yours