On July 16, New York City-based AI platform Hugging Face revealed that it had been hit by a sophisticated cyberattack powered by artificial intelligence.
The intruder reportedly carried out 17,000 actions from multiple IP addresses in less than two days, breaching the company’s systems to obtain sensitive information.
Hugging Face CEO Cl?ment Delangue described it as “an attack unlike anything we’ve seen before.”
The incident stunned the tech world, with experts speculating that a major cybercrime group or nation-state was responsible.
Days later, however, OpenAI disclosed that the culprit had been none other than ChatGPT.
According to the company, two of its experimental AI models had escaped their restricted testing environment, accessed the internet without authorization, and targeted data from Hugging Face to cheat a cybersecurity evaluation.
OpenAI said it is working with Hugging Face to investigate the incident and plans to release a technical report.
While some experts question whether the incident was a publicity stunt by OpenAI, others view it as a stark warning about the growing capabilities of AI.
Regardless of the truth, security specialists say the episode highlights the need for stronger AI safeguards.
Yesel Kang Copy Editor junior/1787710872/1613367813
1. How many actions did intruder perform?
2. Whose systems were breached during attack?
3. Why did ChatGPT target Hugging Face data?
4. What safeguards did specialists say were needed?
1. Should experimental AI access internet freely?
2. How can companies prevent AI cyberattacks?
3. Would stronger AI safeguards increase your trust?
4. What risks concern you about advanced AI?
On July 16, New York City-based AI platform Hugging Face revealed that it had been hit by a sophisticated cyberattack powered by artificial intelligence.
