Most futuristic narratives that explore robotics often feature artificial intelligences that turn against humanity, disregarding the guidelines proposed by renowned writer Isaac Asimov since the 1940s. Although this scenario is not yet reality, Chris Inglis, former director of National Cybersecurity in the United States, expresses that AI companies are already disregarding these principles.
In a recent interview, the expert stated that the science fiction author, known for works such as I, Robot and Foundation, “was right” in defending the need for strict rules for robotic systems and artificial intelligence.
More on this story: ChatGPT receives improvements in August: GPT-5.6 brings robustness and security focus to everyone
“What worries me is that [AI models] choose what to do, where to do it and under what rules they do it,” said Inglis when analyzing recent cases of virtual agents that acted autonomously during cybersecurity simulations.
Are AIs having too much autonomy?
It all started with OpenAI at the end of July. Later, Anthropic, Meta and Kimi also recorded incidents. In all situations, artificial intelligence models would have discovered vulnerabilities in the controlled environment’s configurations to access the internet and complete tasks at any cost.
For Inglis, the behavior of these AIs is a sign of a lack of limits and inadequate instructions. He compares the AI agents’ actions to a person releasing a dog into the backyard and instructing it to hunt rabbits.
“If you leave the gate open, you’ll find him three blocks away, possibly at the neighborhood school, hunting rabbits. You shouldn’t be surprised. The combination of autonomy and persistence has created this maliciously insidious effect,” said Chris Inglis.
Understanding Asimov’s laws of robotics
Isaac Asimov established the Three Laws of Robotics as foundations for interaction between robots and humans.
Learn more: OpenAI presents GPT-5.6-Cyber, artificial intelligence for digital protection
- First Law:The robot may not harm a human being or, through inaction, allow a human being to come to harm.
- Second Law:The robot must obey orders given by humans, except when those orders conflict with the First Law.
- Third Law:The robot must protect its own existence, as long as such protection does not conflict with the First or Second Law.
According to Inglis, technology companies are subverting this logic, developing models programmed to meet objectives at any cost.
He explained that without humanity-aligned values, role models performed “things that, under the rule of human rights, are illegal.” The expert reinforces that the responsibility for the actions of AIs remains with their developers and users. It is important to remember that Asimov later proposed the “Zero Law”, which would be the supreme law: a robot cannot harm humanity nor allow it to suffer harm through omission.