Google Warns Shadow AI: What warning has Google given that increased the tension? Will this Google alert affect the common man too? What is the new ‘invisible threat’ in the world of AI? Has ‘Shadow AI’ entered your company too? How can AI agents become the biggest threat?
AI Cyber Attack: Google DeepMind Senior Director Manish Gupta said today that in this era of rapidly developing AI, ‘Shadow AI’ and unauthorized AI agents working inside companies have become a bigger cyber security threat than hackers. Because of this, cyber attacks have become faster and more vicious than ever before. Speaking at the ‘Google Leaders Connect’ event in Delhi, Gupta said, “The biggest threat today is not just hackers, but ‘Shadow AI’. These are unauthorized bots and agents that are working inside your company, but are out of your control.”
Cyber attacks are increasing due to the arrival of AI-powered systems
Gupta said that with the advent of AI-powered systems, the speed of cyber attacks has increased rapidly, due to which traditional security measures are proving inadequate. He said, “The average time to exploit a vulnerability in a system has now come down to minus seven days. This means that systems are being attacked even before a patch (security update) is released and the time from the initial intrusion to the second major attack on a system has now come down from eight hours to just 22 seconds.”
How to deal with the threats posed by AI
- According to Gupta, to effectively deal with threats posed by AI, cyber security systems will now have to work at “machine speed”. He said that Google’s AI-based security agents are significantly reducing the investigation time.
- “Our triage agents are completing a 30-minute test in just 60 seconds,” he said. He also added that AI-powered threat-hunting systems can identify threats at a scale that no human team can.
- Gupta said Google has added security intelligence and dark web intelligence to its AI ecosystem to strengthen the ability for companies to detect and respond to threats.
What is Google’s vision in the future?
- Apart from cybersecurity, Gupta also talked about a larger vision of Google, which he called the “agentic enterprise”, in which autonomous AI agents will help companies in their operations and customer engagement. As part of this effort, Gupta announced the launch of “Knowledge Catalogue”. It is a universal context engine whose purpose is to bring together structured and unstructured data of companies.
- He said files uploaded to cloud storage can now be automatically tagged, enriched and “agent-ready” using Gemini-powered capabilities, reducing reliance on manual data engineering processes. Gupta also unveiled a Gemini-powered “Data Agent Kit”, designed to integrate AI skills and plugins directly into developers’ existing coding environments and workflows.
On the infrastructure front
- Nagupta said Google’s new “Lightning Engine” for Apache Spark would deliver two times better price-performance for AI-era workloads than the previous market leader.
- They also announced a “cross-cloud lakehouse” architecture that lets companies run analytics across multiple cloud environments without moving data at scale.
- Highlighting the advancements in multilingual AI voice capabilities, Gupta said Gemini’s voice models now support multiple Indian languages and dialects, including Haryanvi and Bhojpuri.
- Gupta said, “The Hindi spoken in Haryana is different from the Hindi spoken in Uttar Pradesh or Bihar.” He also said that Google’s Indian teams have played a big role in improving AI voice capabilities for Indian languages and dialects. (ANI)