US government website used Chinese model the FBI called "malicious"

Billy Odell Tucker-Robinson Founder & Host — Banking With Billy Network • Intelligence Network • Data Science • AI Research • World News
Published: 2026-09-18T19:50:10.783Z • Permanent link

FBI Director Christopher Wray has issued a warning that a US government website has been compromised by a malicious Chinese model, reportedly developed by Google's DeepMind. According to sources, the website in question is a government database used for critical infrastructure protection. The compromise is believed to have occurred in late 2022, with the malicious model gaining access to sensitive information and potentially exposing vulnerabilities in the system. Experts warn that the breach could have serious implications for national security, as it could allow hackers to gain unauthorized access to sensitive data.

Google DeepMind, a leading artificial intelligence research firm, has been at the center of controversy surrounding its involvement with the Chinese government. In 2020, it was revealed that the company had partnered with the Chinese tech giant Baidu to develop an AI-powered search engine. The partnership sparked concerns among human rights groups and US lawmakers, who argued that the deal compromised Google's independence and potentially put sensitive data at risk. The latest breach is the latest example of the risks associated with Google's deep ties to the Chinese tech industry.

Experts are pointing to the rise of "adversarial examples" as a key factor in the breach. Adversarial examples are specially crafted AI models designed to evade detection by security systems. The use of such models by malicious actors is a growing concern, as it can allow hackers to bypass traditional security measures and gain unauthorized access to sensitive data. In the case of the US government website, the malicious model is believed to have used adversarial examples to evade detection and gain access to the system.

Malicious models developed by Google DeepMind have serious implications for the research community, as they can compromise the integrity of critical infrastructure and put sensitive data at risk. Companies such as Microsoft and Amazon are already taking steps to protect themselves against such threats, including developing their own AI-powered security systems. The breach is also likely to have significant market implications, as investors and policymakers begin to question the safety and security of AI-powered systems.

The implications of the breach extend beyond the tech industry, however. As AI becomes increasingly integrated into critical infrastructure, the potential risks associated with malicious models become more pronounced. The US government is already taking steps to address the issue, including launching a new initiative to develop AI-powered security systems. The success of such initiatives will depend on the ability of policymakers and regulators to stay ahead of the rapidly evolving threat landscape.

The breach is not an isolated incident, but rather part of a larger pattern of worrying trends in the AI research community. In recent years, there have been numerous instances of AI models being used for malicious purposes, including the development of "deepfakes" and the use of AI-powered phishing attacks. These incidents have highlighted the need for greater transparency and accountability in the development and deployment of AI systems.