Google threat intelligence: AI-led cyberattacks approaching, quantum encryption breaking by 2029
Google's Threat Intelligence unit warns of impending autonomous AI cyberattacks and predicts quantum computers will break current encryption by 2029, much earlier than previously thought.
📋 Article Processing Timeline
- 📰 Published: April 24, 2026 at 13:46
- 🔍 Collected: April 24, 2026 at 14:01 (15 min after Published)
- 🤖 AI Analyzed: April 25, 2026 at 00:35 (10h 34m after Collected)
Central News Agency
(Central News Agency, reporter Chang Hsin-yu, Las Vegas, 23rd) AI is comprehensively changing the scale, speed, and complexity of cyberattacks. Sandra Joyce, Vice President of Google's "Threat Intelligence" division, stated in an interview today that they have observed Russian military intelligence units connecting to Chinese models with AI malware to generate commands. Currently, human intervention is still involved in such attacks, but it is expected that attacks launched autonomously by AI will happen in the short term.
She also pointed out that the Google team has advanced the timeline for quantum computers to break encryption from 2036 to 2029.
Sandra Joyce has over 25 years of global threat intelligence experience. She joined the team in 2022 when Google acquired the cybersecurity company Mandiant, leading the "Threat Intelligence" division.
Today, she accepted interviews with some Asia-Pacific media at the Google Cloud Next cloud technology conference in Las Vegas, responding to the impact of AI on global cybersecurity, warning that large-scale attacks led by AI agents will emerge shortly.
She noted that the team has observed some preliminary signs, such as Russian military intelligence using AI-driven malware to send requests to Chinese large language models to instantly generate commands. Human intervention is still present, but due to lowered technical barriers, hackers are advancing faster.
The Google team is closely monitoring these attacks. She added: "Organizations that have not yet prepared their system updates or patch management will face huge problems, as these AI tools are becoming increasingly capable at vulnerability scanning."
She said that with the rapid development of AI, the current situation might be completely different in 90 days, or even 6 months.
Speaking on the possibility of quantum computers breaking encryption, she revealed that the team originally expected it to happen in 2036, but has recently moved the timeline forward to 2029, because the speed of technical research progress is quite fast.
She also warned that on this matter, whoever achieves the breakthrough first will be able to decrypt massive amounts of data, and there is not much value in being second.
To combat the threat of malicious attacks, Google announced the launch of three defensive AI agent tools at the cloud technology conference. Joyce stated that while threats continue to expand, defensive capabilities are also improving. The team is now entering a phase of building various agents, with the ultimate goal of managing multiple agent forces to comprehensively understand the entire threat environment. (Editor: Tang Sheng-yang) 1150424
Choose to stand with facts, every sponsorship from you is the power to protect press freedom
Download Central News Agency "First Hand News" APP, grasp the latest news instantly
Text, images, and audio/video on this website cannot be reproduced, broadcast, transmitted, or utilized without authorization.
(Central News Agency, reporter Chang Hsin-yu, Las Vegas, 23rd) AI is comprehensively changing the scale, speed, and complexity of cyberattacks. Sandra Joyce, Vice President of Google's "Threat Intelligence" division, stated in an interview today that they have observed Russian military intelligence units connecting to Chinese models with AI malware to generate commands. Currently, human intervention is still involved in such attacks, but it is expected that attacks launched autonomously by AI will happen in the short term.
She also pointed out that the Google team has advanced the timeline for quantum computers to break encryption from 2036 to 2029.
Sandra Joyce has over 25 years of global threat intelligence experience. She joined the team in 2022 when Google acquired the cybersecurity company Mandiant, leading the "Threat Intelligence" division.
Today, she accepted interviews with some Asia-Pacific media at the Google Cloud Next cloud technology conference in Las Vegas, responding to the impact of AI on global cybersecurity, warning that large-scale attacks led by AI agents will emerge shortly.
She noted that the team has observed some preliminary signs, such as Russian military intelligence using AI-driven malware to send requests to Chinese large language models to instantly generate commands. Human intervention is still present, but due to lowered technical barriers, hackers are advancing faster.
The Google team is closely monitoring these attacks. She added: "Organizations that have not yet prepared their system updates or patch management will face huge problems, as these AI tools are becoming increasingly capable at vulnerability scanning."
She said that with the rapid development of AI, the current situation might be completely different in 90 days, or even 6 months.
Speaking on the possibility of quantum computers breaking encryption, she revealed that the team originally expected it to happen in 2036, but has recently moved the timeline forward to 2029, because the speed of technical research progress is quite fast.
She also warned that on this matter, whoever achieves the breakthrough first will be able to decrypt massive amounts of data, and there is not much value in being second.
To combat the threat of malicious attacks, Google announced the launch of three defensive AI agent tools at the cloud technology conference. Joyce stated that while threats continue to expand, defensive capabilities are also improving. The team is now entering a phase of building various agents, with the ultimate goal of managing multiple agent forces to comprehensively understand the entire threat environment. (Editor: Tang Sheng-yang) 1150424
Choose to stand with facts, every sponsorship from you is the power to protect press freedom
Download Central News Agency "First Hand News" APP, grasp the latest news instantly
Text, images, and audio/video on this website cannot be reproduced, broadcast, transmitted, or utilized without authorization.