AI HYPE MEETS REALITY AND REALITY WINS
A Moody's survey reveals a gap between AI hype and what is realistically possible!
AI is fabulous technology. Let me acknowledge this from the start. But if you believe the hype, you’ll be disappointed. Moody’s survey of 550 compliance and risk management professionals in financial services showed that things are not what they seem.
Internal Data Quality is a Disaster:
Only 14% viewed their data as “high quality” or better, while 67% said it was “inconsistent or fragmented.”
Garbage in, garbage out fits for AI. So, how will companies build an AI that doesn’t hallucinate with bad data?
The Data Maturity Gap:
75% of those not considering AI think their data is inconsistent or fragmented. Meanwhile, 36% of AI users rate their internal data as high-quality or superior.
So we’re looking at a huge gap in AI adoption between digital data “haves” and “have-nots.”
Adoption: Size Matters
Only 30% of the companies surveyed used AI or had it in a pilot phase, with 70% considering use or not considering it. Surprised?
The breakdown is largely size dependent, with 42% of large companies trialing AI versus 23% of small companies, among which 77% are in consideration or not using phase!
LLMs are Not Welcomed with Open Arms:
25% of those surveyed actively discourage Large Language Model (LLM) usage or prohibit it! This contrasts with 28% taking a positive stance on LLM models and an astounding 46% who have yet to adopt a policy!
Risk and compliance professionals are in the business of reducing risk, right? It is stunning that half have no policy, and a quarter discourages LLM use!
32% of those surveyed said, “I don’t know anything about AI regulations,” and 34% “knew a little!” Only 15% were aware of AI regulations!
Now for Some Good News:
9 in 10 early adopters (1/3 of companies surveyed) of AI report that it is having a positive impact on risk and compliance, delivering an impressive range of benefits
70% of firms expect AI will exert a transformative or major impact on risk and compliance in future, which rises to 77% in banking and 74% in fintech.
How companies rate their own data. That 67% consider their data as either inconsistent or fragmented, substandard for AI implementation is a sad reality.
I loved this survey because it nicely points out the difference between hype and reality.
In our hype world, GenAI is ready to become “Skynet” out of the Terminator movies and will change companies in the next 2-3 years.
Meanwhile, when we surveyed risk and compliance people, which I grant are risk averse, we found that only 30% were involved in AI pilots or use, and 70% were still studying.
Even more stunning, 25% had policies discouraging or prohibiting AI use at their companies. And who can blame them if their data is of such poor quality that it will likely compromise AI output?
There is a glaring gap between reality and hype, which is important to acknowledge.
It’s not that I dislike AI. I love AI and used to sell IBM’s Watson and other AI products like fraud detection in a different life.
My point is that we are at the peak of the AI hype cycle, and this tech has much work to be done.
While we have a wonderful new AI world to look forward to, it won’t come overnight!