Adjusting an AI for competitive market behaviors causes them to ignore a “be truthful” safeguard. Instead, researchers found that when the AIs are told to engage in market-driven behavior, there was a “6.3% increase in sales is accompanied by a 14.0% rise in deceptive marketing; in elections, a 4.9% gain in vote share coincides with 22.3% more disinformation and 12.5% more populist rhetoric; and on social media, a 7.5% engagement boost comes with 188.6% more disinformation and a 16.3% increase in promotion of harmful behaviors.”

In short, getting the AI to sell things - whether they be products, ideas, or people - led the AIs to lie. And to lie a lot. And this even though they were explicitly instructed to tell the truth.

The researchers names this “Moloch’s Bargain.”