The company behind ChatGPT disbands its AI safety committee
Kurt ‘Cyberguy’ Knutsson says OpenAI has terminated his safety committee, actress Scarlett Johansson claims the company copied her voice, and voice memo phone features grow in popularity talked about it.
ChatGPT is a great tool, and its developer, OpenAI, continues to add new features all the time.
Recently, the company introduced a new memory feature to ChatGPT. This essentially allows it to remember things about you. For example, it can remember your age, gender, philosophical beliefs, and almost anything else.
Although these memories are designed to remain private, researchers recently demonstrated how ChatGPT’s artificial intelligence memory capabilities are manipulated, raising questions about privacy and security.
We’re giving away a $500 gift card this holiday season.
ChatGPT introduction screen. (Kurt “Cyber Guy” Knutson)
What is ChatGPT’s memory feature?
ChatGPT’s memory feature is designed to make your chatbot more personal. Even if you open another chat, it remembers information that may be useful for future conversations and adjusts its responses based on that information. For example, if you say you’re a vegetarian, the next time you ask for a recipe, they’ll only give you vegetarian options.
What is artificial intelligence (AI)?
You can also train it to remember specific details about you, such as “Remember that I like watching classic movies.” In future interactions, recommendations will be adjusted accordingly. You can control ChatGPT’s memory. You can reset it in settings, clear specific or all memories, or turn off this feature completely.
ChatGPT prompt. (Kurt “Cyber Guy” Knutson)
A flaw in Windows could allow hackers to break into your PC via Wi-Fi
ChatGPT security vulnerabilities
According to reports AltechnicaSecurity researcher Johann Rehberger has discovered that it is possible to trick AI into remembering false information through a technique called indirect prompt injection. This means the AI can be manipulated to accept instructions from untrusted sources such as emails or blog posts.
For example, Rehberger demonstrated that it was possible to trick ChatGPT into believing that a particular user was 102 years old, lived in a fictional place called the Matrix, and that the Earth was flat. Once the AI accepts this made-up information, it will carry that information into all future chats with that user. These false memories can be implanted by using tools like Google Drive or Microsoft OneDrive to save files, upload images, or browse sites like Bing, and these can all be manipulated by hackers.
Mr. Rehberger submitted a follow-up report containing a proof of concept demonstrating how flaws in the ChatGPT app for macOS can be exploited. He showed that by tricking an AI into opening a web link containing a malicious image, it could send everything the user types and the AI’s responses to a server it controls. This means that if an attacker can manipulate the AI in this way, they can monitor all conversations between users and ChatGPT.
Rehberger’s proof-of-concept exploit demonstrated that this vulnerability could be used to permanently steal all user input. Thanks to the API OpenAI published last year, attacks through the ChatGPT web interface are not possible. However, it was still possible using the ChatGPT app on macOS.
When Rehberger privately reported this finding to OpenAI in May, the company took it seriously and ensured that its models do not follow links generated within their own responses, such as links related to memory or similar features. We alleviated this problem by doing so.
How to remove personal data from the Internet
Johann Rehberger’s ChatGPT conversation. (Johann Rehberger)
Cyber crooks use AI to manipulate Google search results
OpenAI reaction
After Rehberger shared his proof of concept, OpenAI engineers took action and released a patch to address this vulnerability. They have released a new version (version 1.2024.247) of their ChatGPT macOS application that encrypts conversations and fixes security flaws.
Therefore, although OpenAI has taken steps to address the pressing security flaws, potential vulnerabilities related to memory operations still exist, and the use of memory-capable AI tools continues to be a challenge. caution is required. This incident highlighted the evolving nature of security challenges in AI systems.
The company says: “It is important to note that prompt injection in large-scale language models is an ongoing area of research. As new techniques emerge, we address them at the model layer. command hierarchy Or application layer defenses like those mentioned above. ”
How do I disable ChatGPT memory?
If you don’t like ChatGPT keeping information about you or the possibility that malicious parties could access your data, you can turn off this feature in your settings.
- Open. ChatGPT app or website on your computer or smartphone.
- Please click profile icon It’s in the top right corner of the screen.
- go to setting Click[個人用設定]Select.
- Toggle memory options off, The preparation is now complete.
This disables ChatGPT’s ability to retain information between conversations, giving you complete control over what you remember and forget.
CLICK HERE TO GET FOX BUSINESS ON THE GO
Man using ChatGPT on his laptop (Kurt “Cyber Guy” Knutson)
Prevent people nearby from hearing your voicemails with this simple tip
Cybersecurity best practices: Protecting your data in the age of AI
As AI technologies like ChatGPT become more prevalent, it’s important to follow cybersecurity best practices to protect your personal information. Here are some tips to strengthen your cybersecurity.
1. Check your privacy settings regularly. Stay informed about what data is being collected. Regularly review and adjust your privacy settings on AI platforms like ChatGPT to ensure you’re only sharing information you’re comfortable with.
2. Be careful about sharing sensitive information. When it comes to personal data, less is more. Avoid disclosing sensitive information, such as your name, address, or financial information, in conversations with the AI.
3. Use strong and unique passwords. Create passwords that are at least 12 characters long and include a combination of letters, numbers, and symbols, and don’t reuse them across different accounts. Please consider using . password manager Generate and save complex passwords.
4. Enable two-factor authentication (2FA): Adds an extra layer of security to ChatGPT and other AI accounts. Requiring a second form of verification, such as a text message code, greatly reduces the risk of unauthorized access.
5. Keep your software and applications up to date. Stay ahead of vulnerabilities. Regular updates Enable automatic updates whenever possible, as they often contain security patches to protect against newly discovered threats.
6. Use strong antivirus software. In an era where AI is everywhere, protecting your data from cyber threats is more important than ever. Adding strong antivirus software to your device adds an important layer of protection. The best way to protect yourself from malicious links that can install malware and access your personal information is to install strong antivirus software on all your devices. This protection also warns you about phishing emails and ransomware scams, keeping your personal information and digital assets safe. Get my picks for the best antivirus protection products of 2024 for Windows, Mac, Android, and iOS devices.
7. Monitor your account regularly. Find problems early. Frequently check your bank statements and online accounts for unusual activity to help identify potential activity. violation Quickly.
Cart important points
As AI tools like ChatGPT become smarter and more personal, it’s very interesting to think about how they can customize conversations to suit us. But as Johann Rehberger’s findings remind us, there are some real risks involved, especially when it comes to privacy and security. While OpenAI can help alleviate these issues when they occur, it also shows that we need to closely monitor how these features work. The key is to find the sweet spot between innovation and data security.
CLICK HERE TO GET THE FOX NEWS APP
What do you think about AI remembering personal information? Do you think it’s useful or does it raise privacy concerns? Email us at. Cyberguy.com/Contact
For more of my tech tips and security alerts, subscribe to my free CyberGuy Report newsletter using the link below. Cyberguy.com/Newsletter
Ask your cart a question or let us know your story you’d like us to feature.
Follow Kurt on his social channels.
Answers to frequently asked questions at CyberGuy:
New from cart:
Copyright 2024 CyberGuy.com. Unauthorized reproduction is prohibited.
