-
22,424
Abibisika (Black Gold) Points
Be aware of the PROS and CONS of AI make believe free and convenience values.
Is Chat-GPT Safe? Cyber-security Concerns and Top Tips
Released in November 2022, Chat-GPT is an advanced AI chat-bot developed by Open-AI, a startup founded in 2015. Chat-GPT is what’s known as a natural language processing tool, based around a powerful AI that uses data from the internet to compose its own text. It can answer almost any question and produce almost any content. In its own words:
“Chat-GPT is a cutting-edge conversational AI model that leverages the power of Generative Pre-trained Transformers (GPT) to understand and generate human-like text. Whether it’s answering questions, providing explanations, generating creative content, or assisting with various tasks, Chat-GPT aims to simulate human-like interaction, offering users informative and engaging experiences.”
Is Chat-GPT Safe?
Any new technology raises concerns about cyber-security, and that’s the case with Chat-GPT, too. While it can do many great things, the chat-bot can also be used to write malicious coding and create viruses, spread misinformation, and produce copy for phishing materials. Then there is the issue of data security. Read on for the low-down.
Malicious AI-Generated Content
Researchers have warned that cyber-criminals can use Chat-GPT to compose the text for phishing emails, texts, and call scripts — meaning more phishing and more cyber threats. The generated text is so effective that Open-AI withdrew a tool it had designed to spot AI-generated content, due to poor accuracy results. Nonetheless, Open-AI offers three suggestions for spotting AI-generated text:
Check for patterns and inconsistencies: AI-generated content may have patterns or inconsistencies that indicate it was generated by a machine, such as repeating the same phrase multiple times or using the same structure repeatedly.
Look for signs of human error: AI-generated content may contain errors or inconsistencies that would not be present in human-generated content, such as awkward phrasing or lack of context.
Check the context: AI-generated content may be out of context or not make sense in the context of the conversation.
This guidance is by no means foolproof, but it’s sound advice, nonetheless. In simple terms, take your time and be skeptical!
Open-AI’s Privacy Policy
Open-AI’s privacy policy outlines the types of data collected: including contact details, date of birth, transaction history, network activity, content, location, and log-in credentials. While some data sharing is necessary the potential for misuse and leaks remains a valid concern.
In addition, Chat-GPT conversations are not confidential. While Open-AI has protections in place to prevent sharing personal information, there’s still a risk of exposure. Your chats may be used to train future versions of the model, for example, while all chats are stored for 30 days to monitor for abuse.
Accidents and Data Leaks
In March 2023, Open-AI encountered an issue where user chat history got mixed up. Some Chat-GPT users saw other people’s conversation history instead of their own. There were also concerns that financial information from Chat-GPT-Plus subscribers might have been leaked. To their credit, Open-AI did quickly address the bug and subsequently published a report — however, leaks and breaches remain a risk inherent to any online service.
Misinformation
Fake news websites (often referred to as news-bots) are leveraging AI to create news articles — although, due to the nature of AI, the accuracy of such “news websites” cannot be trusted. These sites are designed to generate ad revenue for the site owners, and because AI can create content so quickly, the sites can publish many AI-generated articles in a short amount of time, each containing lots of ads.
Fake Chat-GPT Apps and Websites
In previous investigations, we also found large numbers of fake Chat-GPT apps and phishing web-pages, some containing malicious Trojan viruses and others being spam-generating PUAs (“potentially unwanted app”).
Protecting Your Identity and Personal Info
Head over to our new ID Protection platform, which has been designed to meet the security and privacy threats we now all face.
