ChatGPT, OpenAI’s viral chatbot, is experiencing outages caused by a targeted attack, according to an OpenAI spokesperson. As of Thursday, users were still receiving some error messages.

More than 92% of Fortune 500 companies use the platform, up from 80% in August, and they span across industries such as financial services, legal applications and education, OpenAI Chief Technology Officer Mira Murati told reporters Monday.

“We are dealing with periodic outages due to an abnormal traffic pattern reflective of a DDoS attack,” OpenAI shared on Wednesday evening, just before 11 p.m. ET. “We are continuing work to mitigate this.”

The company said customer information had not been compromised.

The issues follow Wednesday’s outage across the company’s tools and services, including the API used by more than two million developers, and seemed to begin Wednesday morning, when OpenAI reported errors “impacting all services,” according to its status page. ChatGPT users were told “ChatGPT is at capacity right now,” but after implementing a fix, things returned to normal, per OpenAI — but not for long.

The attack also follows OpenAI’s first in-person event on Monday, where the company announced it had surpassed 100 million weekly active users. It also announced its most powerful artificial intelligence model yet, GPT-4 Turbo, and a new option allowing users to create customized versions of ChatGPT.