An insider at OpenAI says the company’s security teams are struggling to keep up as AI‑related incidents climb sharply. The employee, who asked to remain anonymous, pointed to a recent spike in model jailbreaks, data leaks, and malicious prompt injections that have strained limited resources and forced overtime.
According to the source, the surge is driven by the rapid rollout of powerful language models that lower the barrier for misuse. Security analysts now spend most of their day reacting to alerts rather than building preventative defenses, leading to burnout. The team reports a 60 percent increase in incident tickets over the past six months.
OpenAI acknowledges the challenges and says it is hiring additional engineers and investing in automated monitoring tools. The company also plans tighter integration with industry‑wide threat sharing groups to anticipate emerging risks. Observers warn that without swift action, the tension between rapid innovation and security will only intensify.
Source: Read original article