Sometimes AI seems pretty impressive — at least in its ability to whip up several images of a cat playing a banjo in a matter of seconds.
But is it always? SEC Chair Gary Gensler recently brought up the concept of “AI washing” in a pop culture-laden speech, comparing it to The Music Man, a musical about a big-talking con man.
AI washing is when companies create false buzz and manipulate investors by:
- Being misleading about the capabilities, limitations, or risks of their AI products.
- Lying about how and when they’re using AI.
Why is this happening?
Everyone wants to be part of the hot new thing. Goldman Sachs found that a record 36% of S&P 500 companies talked about AI in their Q4 earnings reports, per Business Insider.
But many companies are nebulous at best about what AI can do and how it will integrate into existing operations, while others are hyping a future that may never occur.
Example: I received a press release about an AI girlfriend who doesn’t nag you to do the dishes. She also doesn’t really exist, and unless she’s a WiFi-enabled dishwasher, you’ll eventually still have to do your own dishes.
Does AI washing have consequences?
Yes. Hype too hard, and the SEC will come for you. It charged two companies earlier this month.
- Toronto-based investment adviser Delphia claimed it used AI to predict upcoming companies and trends when it didn’t have those capabilities. It agreed to pay a $225k fine.
- Bay Area-based Global Predictions falsely claimed it was the “first regulated AI financial advisor” and that it offered “AI-driven forecasts.” It agreed to pay a $175k fine.
While AI does have great potential, companies should also be realistic about what it can do.
One company found its customer support agents were 14% more productive when integrating AI, though its higher-skilled agents saw little benefit.
And companies that have attempted to use AI to fully create articles, art, or other creative work have had mixed results. May we never forget that Glasgow Wonka experience.