From toothbrushes that sense your general oral health to refrigerators that learn how much milk is left for your next cup of coffee, AI is everywhere. It’s the hot topic buzzword that’s working its way into every sector of society, from the bathroom to the boardroom. It’s become an important component in launching innovation–and winning over customers who are looking for an easy answer to why one product is “better” than another.

Tay AI

Salesforce made headlines today for launching its AI feature Einstein in its sales software on the same day that Oracle unveiled Oracle AI, and both products are aimed at providing business customers with real-time data that not only tracks your workflow, but also makes predictions (hence the “intelligence” part of AI). And the accompanying buzzwords like “machine learning” are impressive, to say the least. But there are some questions that are still waiting to be answered.

The first–as always–is to ask if this is really a game changer, or is it just a fancy whirligig intended to amaze the natives. Does your toothbrush really need to track your brushing habits and then adjust itself based on that data? And do consumers actually intend to make a lifelong habit of downloading their data and studying it, looking for trends that correlate to the number of cavities they have? Hint: dentists would be thrilled to see regular brushing and flossing from their patients, so you might start there.

But the bigger implication in the rush to incorporate AI into every aspect of consumer life is the security implications. Is AI going too far, too fast for security to keep up with it? It’s already a certainty that the legal system hasn’t kept up, as legislators scramble to create laws that protect IoT users, for example. And even stripping away the malicious criminal intent behind security flaws, are human users savvy enough to be handed AI devices without putting themselves at risk? The infamous Google Nest hacking in 2014 might indicate otherwise.

If the announcements and the unveilings are teaching really teaching us anything, it’s that the race to market might not be the safest approach. One surefire way to lose customers will be to provide a product that compromises itself in its attempt to learn from them.