AI Models at Risk? Eric Schmidt on the Threat of Reverse Engineering

Published: 13 Oct 2025

Author: Precedence Research

Share :

Amid the ongoing wave of AI and its unprecedented capabilities, a former CEO of Google, who led the Company for a decade in the early 2000s, made a statement during a conference about AI, saying that AI can be reverse-engineered and there is evidence for it. So, it would be catastrophic if AI models fell into the wrong hands. The former CEO of Google, Mr. Eric Schmidt, has sounded an alarm over the implications of artificial intelligence on who is behind its fundamental building and algorithms.

He said at the Sifted Summit tech conference in London, “There’s evidence that you can take models, closed or open, and hack them to remove their guardrails. During training, they learn many things. A bad example would be that they learn how to kill someone.”

While AI models continue to generate buzz in the technology world and improve in tasks like coding, reasoning, and bug detection, they are also highly susceptible to hacking and jailbreak attacks by malicious entities. This is an alarming aspect of the potential threat posed by AI and its misuse, which everyone seems to overlook. There is strong evidence of its potential for reverse engineering. In fact, experts fear that AI models to become potential cybersecurity weapons.

As generative AI adapts itself through data and learning automatically, it poses a high risk of malware attack that can spread rapidly across the codebase, unveiling a torrent of malicious code at the push of a single switch if proper safety protocols aren't in place.

It's not the first time that tech leaders have warned people about potential threats from AI and its rapid integration into nearly every sector alongside other emerging technologies. A few years ago, popularly known as” Godfather of AI”, Mr. Geoffrey Hinton also pointed out risks involved with the developing of AI models like ChatGPT.

He further mentioned in an interview with the BBC in 2023 that, “Right now, they’re not more intelligent than us, as far as I can tell, but I think they soon may be. I have concluded that the kind of intelligence we’re developing is very different from the intelligence we have. We are biological systems, and these are digital systems. The big difference is that with digital systems, you can have many copies of the same set of weights, the same model of the world.” He further said that AI might set its own agenda, such as “I need more power to work efficiently,” which could threaten humanity.