Google Concerned About Increasing Deployment of Agentic AI for Cyberattack Operations

Google has warned of the dangers of agentic AI in cyberattacks, warning that advanced AI could make cyberattacks more dangerous and harder to detect. The announcement comes as cyber risks become more severe as cybercriminals seek to use autonomous AI agents to take on every stage of the cyberattack chain. This refers to an AI…

Google has warned of the dangers of agentic AI in cyberattacks, warning that advanced AI could make cyberattacks more dangerous and harder to detect. The announcement comes as cyber risks become more severe as cybercriminals seek to use autonomous AI agents to take on every stage of the cyberattack chain.

This refers to an AI system capable of ‘doing things’. An agentic system makes a multi-step decision and then executes that decision with limited human intervention – it observes that it is better to act in a way rather than be told what to do, in the first place.

While threat actors are exploring how AI can be utilised to aid reconnaissance, social engineering, and investigations of vulnerabilities, and the orchestration of malicious campaigns, Google cybersecurity teams have said. They are also utilising AI-generated phishing attacks, which means email messages or other forms of communication are automated so the deception doesn’t appear too different from real contact.

One of the greatest fears of cybersecurity professionals is the ongoing proliferation of advanced AI tools. Cyber criminals would be able to use automation to more quickly and cheaply perform sophisticated operations and execute a greater number of attacks on corporations, governments and the general consumer.

The Google warning also highlights the need for us to secure our digital infrastructure in the future as artificial intelligence advances. Researchers advocate for continuous supervision, better authentication, incremental software updates, and greater employee awareness.

Security controls will be more essential as AI is adopted more widely at work. The emergence of agentic AI in cyberattacks reminds us that innovation is only responsible innovation if it is watched, monitored and protected against abuse.

​

Leave a Reply

Your email address will not be published. Required fields are marked *