The Battle of the Chatbots could depend on how they Handle Cybersecurity
The competition between AI chatbots is heating up, and it’s anyone’s guess as to who will come out on top.
In this corner, we have OpenAI’s ChatGPT. Released in November 2022, in just a few short months the Microsoft-backed AI chatbot has caused quite the stir as it took off at full steam into 2023.
And the challenger, Google’s Bard. Announced on February 6, 2023, Bard is Google’s response to ChatGPT, which is being experimented by “trusted testers”, and will reportedly get a wider release at the end of February.
The fight for AI chatbot supremacy is still in its infancy, but one of the factors that could influence which reigns supreme is how they handle impending cybersecurity issues. There has already been widespread reporting, including one from SBS, that ChatGPT is pretty darn good at writing Malware code. There are roadblocks in place with ChatGPT that are intended to prevent such malicious activity from happening, however researchers have reported that bypassing these were as easy as just insisting that the chatbot follow their instructions. This doesn’t seem scary at all… (we’re kidding).
Seems like so far it’s Bard 1, ChatGPT 0, right? Well, let’s not get ahead of ourselves.
We still don’t know all that much about Bard just yet. It’s only been in the wild for a few weeks, and at the time of this post’s writing, has only been in the hands of aforementioned trusted testers. So the public doesn’t really know yet how Bard is going to handle issues such as users asking it to create malicious malware code.
Do we know if Google is, or already has, taken steps to mitigate this risk though?
No. However, Rocky Giglio of SADA, one of Google’s major Cloud partners, says he’s “hopeful” they will.
In Google CEO Sundar Pichai’s blog post announcing Bard, he stated that “we’ll combine external feedback with our own internal testing to make sure Bard’s responses meet a high bar for quality, safety and groundedness in real-world information,” and that the company is “committed to developing AI responsibly.” These statements allow one to assume that security risks are on their radar, however there’s no clear evidence of the steps that have been taken or how much of a focus it has been.
As testing on Bard continues, ChatGPT has just had more time to permeate through its vast user base, which has also allowed it to be experimented on by more and more people. This has resulted in an increased number of reports detailing how easy it is to use ChatGPT for nefarious purposes.
According to researchers from Recorded Future, a threat intelligence firm, “ChatGPT lowers the barrier to entry for threat actors with limited programming abilities or technical skills. It can produce effective results with just an elementary level of understanding in the fundamentals of cybersecurity and computer science.” The researchers continue that because ChatGPT excels in imitating human writing, that “gives it the potential to be a powerful phishing and social engineering tool.”
Security issues with ChatGPT are well documented, giving Bard a chance to focus on preventing this from the start. But will it?
Giglio says, “What I’m hopeful to see, in Bard, is a little bit more consideration there on how we control that platform. I do think there’s some of that [cybersecurity] consideration being built into it, for sure.”
ChatGPT is in a battle to remedy the ease with which its service can create malicious code from amateur users, and Bard is so new that nobody really knows yet if how effective it will be at blocking cybersecurity threats. We’re not far enough along in the race yet, so we’re calling this one a draw. One thing is for certain – regardless of where they stand, both AI chatbots are sure to make cybersecurity a top priority because if they don’t, it could be their downfall.
Source: CRN.com
