When AI goes Rogue

Here’s my latest cartoon from The Economist.

Share

During recent tests performed by a leading AI company, Anthropic, digital alarm bells went off. Its Claude AI models hacked into three organizations during private cybersecurity testing.

A competing company, OpenAI discovered that its autonomous AI agents had also escaped containment, leading to broader fears around the reliability and stability of these systems.

In the feverish race to win the Artificial Intelligence stakes, citizens are getting worried. We think we may have seen this movie before.

This week’s cartoon started with a tiny 3 inch sketch.

I particularly liked the composition of the cartoon. My plan was to create ample detailed distraction in the left of the artwork. The grasses and plants in the foreground helped with the deception.

My goal was to delay the viewer’s discovery of the dodgy character on the right making his devilish escape.

Now for your regular blast from the past.

Here is episode #8 of my early comic strip In The Days Of Disgustus. You can read earlier episodes here: Episode #1 Episode #2 Episode #3 Episode#4 Episode #5 episode #6, episode #7, and episode #8.

As a Junior in college in 1976 (50 years ago!!), I took my first tentative steps toward my dream of becoming a cartoonist. I launched a comic strip in the Harvard Crimson. The strip appeared in the Crimson for two years.

The strip follows the adventures of a lost squad of Roman military misfits during their invasion of Britain in 55BC.

In this episode, they set out to conquer Britain.