AI training startup faces data breach, lawsuits, and lost contracts after its tool was hacked with malware
San Francisco: Mercor is a big tech company that trains computers to think. In March, someone broke into Mercor’s computer systems.
The hackers stole lots of important things like people’s names and jobs, computer code, and special passwords. Mercor said they are trying to fix the problem and tell their workers what happened.
This happened because another computer tool called LiteLLM was hacked for about 40 minutes. That tool helps people use different types of computer brains. During those 40 minutes, bad software was hidden inside it that could steal passwords.
Some big companies like Meta stopped working with Mercor because of this. Meta is very careful about keeping its secrets safe. Other big tech companies like OpenAI are checking if they were hurt too, but they haven’t stopped working with Mercor yet.
Some workers who used to work for Mercor are mad because their personal information was taken. They are taking Mercor to court about it. One court case is even talking about LiteLLM and a security company called Delve.
People say Delve may have lied about its safety tests. Even though Delve says this isn’t true, many companies don’t trust them anymore. LiteLLM stopped working with Delve and found a new safety company.
Mercor was doing really well before this happened. They were going to make more than one billion dollars this year. Now it’s not clear if the company will keep growing or if this problem will hurt them a lot.