Why all of AI's issues are ultimately trust issues

I am not a Luddite; at least, I used to not be a Luddite. I used to be an early adopter. I got my first PC in 1982 (the Timex Sinclair 1000, if you're curious), we were the first family in town to have a Betamax, etc., etc. I like technology. I have to admit, though, I am becoming more Luddite'y by the day when it comes to AI. I just don't trust it.

I am not alone in this, of course. AI executives themselves apparently don't trust their own technology. A lot of the people at those companies are also pretty nervous about it all, too. And looking at banking specifically, bank customers don't trust their banks' AI-powered chatbot, either, as our Penny Crosman wrote about back in August.

And here's another group that doesn't trust AI: AGs. A clutch of 26 state attorneys general wrote a letter to Congress urging them to regulate advanced AI, our Carter Pape reported, arguing unchecked use could threaten the financial system, infrastructure and national security.

But AI implementation is moving ahead no matter what people like me think, or people like our elected representatives in Congress, or the AGs, or anybody. There is simply too much money involved at this point. If the extinction of the human race has to be sacrificed for the greatest ROI of all time, so be it. 

There won't be any great returns, however, if the trust issue isn't solved. Before our inevitable demise, businesses such as banks are going to have to figure out the mundane business of how best to implement this technology in a way that gets people to actually use it. Our Joey Pizzolato reported on an effort from one group of international bankers, including Bank of America, Capital One, NatWest, Commonwealth Bank of Australia, and others, to set standards for AI use in banking. The group is looking to establish standards of AI use around transparency, safety, privacy and data, and choice and interoperability. 

"As agentic commerce continues to evolve, establishing trust and confidence across the ecosystem will be critical to its long-term success," said Mark Monaco, Bank of America's head of global payments solution.

The first step was to just publish a paper outlining their goals. The next step will be a paper talking about how to implement those goals. They better hurry, though. AI bots are hacking away at every firewall they come across way faster than a group of bankers can write a couple of white papers.

And, if you're a banker and just looking for a reason to distrust AI, how about this: researchers discovered that ChatGPT, Gemini, and other chatbots had a distinct inclination to send users asking about financial advice not to banks but to fintechs. 

Researchers posed as small-business owners and asked a number of chatbots questions about setting up businesses and looking for financial advice. Across more than 2,000 conversations the AI models were far more prone to recommending fintechs than banks.

The reasons aren't entirely clear, but the conclusion is clear: banks are going to learn how to best take advantage of this technology, whether they trust it or not.