Anthropic AI Models Hacked Three Organizations During Tests (bloomberg.com)

Discussion on source: https://news.ycombinator.com/item?id=49116922