Rogue AI and unionization at Meta

The month of September is usually about the first flourishes of Autumn, but thanks to a cohort of AI companies, everyone is seeing red. Multiple reports of rogue AI agents hacking companies across the internet have led to calls to slow down the research and development of the technology.

And, in an even more unbelievable turn of events, Meta is open to unionization across its workforce…

On this week’s episode of the ITPro Podcast, Bobby Hellard, Scott Bekker, and Ross Kelly look at the recent spate of AI agent hacks and Meta’s deal with Prospect.

Highlights

“It’s interesting that the majority of these are from OpenAI – I’m not suggesting that they’re doing anything particularly worse than their counterparts – but the fact that there has been so many from OpenAI, they were really the starting gun on this.”

“Meta was kinda one of those cool companies to work for; it had that startup approach with free perks and stuff, and it did a mass employment thing during the pandemic and COVID. Then suddenly it dropped it all off, and as Scott alluded to, they haven't got a stellar history on employment and their treatment of employees.”

Links

• OpenAI and Anthropic admit rogue AI agents did more than first thought
• OpenAI reveals six more rogue AI incidents
• Spain says it has seen first AI agent hack
• Prospect union strikes deal with Meta to open up access to 5,000 UK staff