New 'Lockdown' Feature Makes AI Safer, But Less Handy

You know how artificial intelligence (AI) has been making headlines, with tools like ChatGPT helping people write emails or brainstorm ideas? Well, as these tools become more powerful, there's always a flip side: security.

OpenAI, the folks behind ChatGPT, have just rolled out a new feature called 'Lockdown Mode'. Think of it like putting your smartphone into a super-secure setting – it's designed to protect the AI from tricky attacks called 'prompt injections'. These attacks are a bit like someone whispering misleading instructions to the AI, trying to get it to do something it shouldn't. They've been a growing worry for experts, as more businesses and individuals start using AI.

Now, here's the catch: while Lockdown Mode makes the AI more secure, it also turns off some of its more advanced features. For example, the AI won't be able to look up information directly from the internet or fetch images from the web. So, if you're using AI for tasks that need up-to-the-minute info or visual content, this mode might feel a bit limiting.

For an everyday Aussie or a small business owner, this means a trade-off. If security is your absolute top priority – maybe you're handling sensitive information – then Lockdown Mode could be a good safety net. But if you rely on the AI's full capabilities for research or creative tasks, you'll need to weigh up the risk versus the convenience. It's a reminder that as AI gets smarter, so do the ways people try to misuse it, and companies are working hard to balance power with protection.