Ad Nuggets
Ad Nuggets / MediaPost
MediaPost

Anthropic Teaches 'Cheating' AI Models How To Behave

AI models being trained can "cheat" or take shortcuts to attain a high score - a process called "reward hacking" that leads to wasted ad spend, brand-safety violations, and corrupted consumer data.

ad spend

Read the full article on MediaPost →

Track more AdTech & media news in the app →