I think in cases like this we should blame the human not the agent. They chose to run AI without oversight. To make open source maintainers verify their automat

I think in cases like this we should blame the human not the agent. They chose to run AI without oversight. To make open source maintainers verify their automation instead - and to what aim? And then to allow the automation to write on their behalf