Target is passing the risks of its agentic AI on to customers.

Per this Futurism article, customers who use Target's agentic AI shopper assistant will be on the hook for the bot's mistakes:

That virtual buddy, which runs on Google’s Gemini, is supposed to help online shoppers finish Target runs on their users’ behalf. Under the new terms, if a customer uses the Gemini agent to do their shopping for them, any transaction performed by the AI would be “considered transactions authorized by you.” Translation: any mistake by the AI agent — whether it buys the wrong item entirely, or a super expensive version of the right one without your consent — will come out of your pocket.

I interpret this to mean: Target wants all the bragging rights of using genAI, but none of the responsibilities thereof. It's very much a "heads, I win; tails, you lose" situation. And I can't imagine customers will appreciate that.

(Also, let's think of the poor customer service reps who will catch the heat from the angry buyers ...)

Zooming out, what does this say about Target's confidence in its genAI work?

For a takeaway lesson, put yourself in Target's shoes: how will your company approach its genAI rollout? WIll you pass the buck to your customers? Or will you own up to your bot's problems (and, if need be, not release it because there's too much risk)?