When AI security researchers take a look at your flagship AI product and note that there are “virtually unlimited” ways to bypass its security, that’s seldom a good sign. And that’s exactly what happened to Alphabet (NASDAQ:GOOG) (NASDAQ:GOOGL) as Google’s AI product, Bard, came under fire from researchers. Alphabet has so far lost 2.5% in Wednesday afternoon’s trading.
A coalition of AI researchers from Pittsburgh’s Carnegie Mellon University and San Francisco’s Center for AI Safety noted that there are “virtually unlimited ways to bypass the content moderation” function of not only Bard, but also ChatGPT and even Anthropic’s Claude system. Moreover, none of the researchers—or anyone else, for that matter—could figure out a way to prevent such activity from taking place. With the safety systems so readily removed, the researchers cautioned, AI systems could be used to generate “hateful content” or “advise on illegal activities.”
While the term “hateful content” has been somewhat stretched in recent years, one thing is certain: no one wants to be accused of generating it. Ever. Worse, this comes at a time when Google is firing up its AI ambitions. Recent announcements note that Google Assistant will get an AI makeover. The end goal of that result: to “supercharge” the tool “like ChatGPT.” The combination of superpowered AI whose ethical boundaries may be, at least, somewhat suspect isn’t a good combination. It’s also likely not particularly attractive to investors, either.
However, analysts aren’t falling away just yet. Currently, Alphabet is a Strong Buy, thanks to six Buy ratings and two Hold. Meanwhile, thanks to an average price target of $142.63, Alphabet stock offers investors a much more attractive 10.9% upside potential.
