Deepseek’s AI model proves easy to jailbreak – and worse

/in

In one security firm’s test, the chatbot alluded to using OpenAI’s training data.

Latest stories for ZDNET in Security – ​Read More

Regional healthcare systems report data breaches affecting more than 1.5 mi...Regional healthcare systems report data breaches affecting more than 1.5 millionState Data Privacy Regulators Are Coming. What Story Will You Tell Them?State Data Privacy Regulators Are Coming. What Story Will You Tell Them?