By A.J. Vicens and Sam Tapahreti

May 11 (Reuters) - Alphabet Inc.'s Google said on Monday that hackers from a prominent cybercrime group used artificial intelligence to discover a previously unknown software vulnerability and were able to exploit it for the first time.

Google stated in a report issued by its Threat Intelligence Group that the planned attack targeted a widely used open-source system management tool, but was blocked before it could be exploited in a "mass exploitation incident".

This incident marks the first time Google has identified hackers using artificial intelligence to discover new vulnerabilities and attempt to exploit them on a large scale.

John Hultkost, the group's chief analyst, said the findings likely represent what he called "the tip of the iceberg" regarding how criminals and state-backed hackers are developing innovative hacking and penetration techniques using artificial intelligence.

The report explained how hackers delegate parts of their online operations to artificial intelligence, exploiting it to independently search for software vulnerabilities and help develop malicious software.

The researchers said this shift represents an early step toward more autonomous online cyber operations, at a time when hackers are beginning to rely on artificial intelligence systems, not only as research tools but also as active components that can analyze targets, create codes and symbols, and make decisions with limited human supervision.

These findings come at a time when governments around the world are struggling with how to regulate powerful new AI models that could make it easier for hackers to identify targets and launch attacks using known and newly discovered software flaws.

These findings reflect recent warnings from financial regulators in Europe that rapidly evolving AI models are increasing the speed and scope of cyber risks at a time of escalating geopolitical tensions.

The report stated that cyber intruders, as well as hacking and penetration groups linked to countries such as China, Russia and North Korea, are already experimenting with integrating artificial intelligence directly into the flow and sequence of tasks related to online cyberattack operations.

Although these technologies are still in their early stages, Google has warned that they could accelerate cyberattack campaigns by reducing the time and expertise needed to launch complex attacks.