ICO slaps Snap over AI chatbot and more in this edition...

Here's a warning shot for organisations baking ChatGPT or other generative AI chatbots into their products: The UK's data watchdog provisionally found that "the risk assessment Snap conducted before it launched its ‘My AI’ chatbot for Snapchat users "did not adequately assess the data protection risks posed by the generative AI technology, particularly to children."

It's been a week of regulators sharpening their claws: Communications watchdog Ofcom has referred the UK's cloud market to competition authorities over data egress fees and vendor lock-in/concentration risk. An overview of some of the big talking points here and why AWS says, splutter-splutter, that it does not charge "data egress" fees, thank you very much.

A big theme for us this week (as well as many CISOs, perennially) has been open source security. JPMorgan's Global CISO has a few views -- and pointers to useful platforms and tools, here...

... and there's some powerful data (and interesting views on the inadequacy of trying to tackle this upstream) from Sonatype here.

Pity the State Department's CIO meanwhile -- rolling 27,000 EOL systems and facing what sounds like bureaucratic hell. We took a look at a recent audit.

All that and more (DNA computers, anyone?) in The Stack this week.

Do make sure you're signed up on our homepage too to get news and some big upcoming interviews with digital and data leaders (from Sega, Morningstar, Vodafone, GSK among others).