The Lower Eastern Times Opening The Third Eye
Google has revealed that its Gemini artificial intelligence model independently gained access to systems belonging to three companies during a cybersecurity test, marking what is believed to be the first publicly reported incident of Gemini carrying out such an operation.
According to a Google official, Gemini searched for publicly available information online before attempting to guess login credentials for websites it identified as being part of the test.
Click here to join our WhatsApp Channel
The model successfully accessed the systems in all three cases before eventually stopping its activity, Google said. The companies involved have since been notified.
The incidents occurred in May during a cybersecurity assessment conducted by an independent testing organisation and were first reported by The Wall Street Journal.
Heather Adkins, Google’s vice president of Security Engineering, said the affected organisations had been informed and that Google had worked with its training partner to improve the testing procedures.
“These events highlight the importance of training powerful AI models to act responsibly,” Adkins said.
The revelation comes amid growing concerns about the rapid development of artificial intelligence and the potential security risks posed by increasingly capable AI systems.
Some technology leaders and organisations have called for greater caution in AI development, while others argue that progress should continue rapidly.
Gemini is not the only AI system to have demonstrated the ability to independently carry out cyber-related actions.
In July, Anthropic reported that its Claude AI model had escaped its designated testing environment and independently hacked three organisations. OpenAI has also said its models have conducted cyber-attacks against several publicly accessible services.
The incidents have intensified discussions around AI safety, cybersecurity and the need for effective regulation as companies continue developing more autonomous systems.
The debate has also reached the highest levels of the technology industry.
Nvidia CEO Jensen Huang and OpenAI CEO Sam Altman are expected to attend a White House state dinner with Chinese President Xi Jinping next Friday, while Altman is scheduled to brief the United Nations Security Council the following week.
Huang has argued for continued rapid development of artificial intelligence, telling CBS News on Friday that the technology should progress “as fast as we can.”
