Asimov’s laws are not sufficient to ensure the safety of robotics and artificial intelligence

Most science fiction fans know that Isaac Asimov proposed his Three Laws of Robotics in the March 1942 issue of Amazing science fiction magazine and later republished them in his famous 1950 collection, I, Robot. Since then, readers have pondered both the obvious and less obvious implications of his dictates. In an era when people fear that artificial intelligence threatens all of human civilization, Asimov’s observations about the dangers of robots to humans seem even more prescient.

Asimov’s three laws of robotics are as follows:

1. A robot cannot harm a human being or allow, through inaction, a human being to be harmed.
2. A robot must obey orders given by humans, except where such orders conflict with the First Law.
3. A robot must protect its existence as long as such protection does not conflict with the First or Second Law.

When thinking about artificial intelligence, it is tempting to focus on the first and best-known of the three laws. Surprisingly, only one corporate entity has adopted this principle as a core element of its software development: Advanced Robotic Technologies Inc., which specializes in generative artificial intelligence (gen AI) applied to the production of large-scale industrial products.

To any attentive observer it would seem that such a fundamental principle could be incorporated into the core algorithms of every AI company in the world. Alas, this is not the case. One might reasonably assume, as many people do today, that without such prohibitive instructions built into a robot’s basic “intelligence,” increasingly capable robots could eventually pose a significant risk to human safety.

Given the controversy that has recently arisen over this very topic, it would be worth reflecting on the warning we just received from Dario Amodei, CEO of Anthropic, on September 12 in his post titled “We must patrol the border.” In it, he explicitly calls for slowing down the development of frontier AI models in general, and agentic AI models in particular.

Amodei expressed concern that swarms of autonomous AI agents can now perform complex tasks, act in unison with each other without any human knowing, and even cover their tracks when they violate basic security principles.

Tech leaders worry about the social risks of artificial intelligence

Sam Altman, CEO of Anthropic’s competitor OpenAI, said he agreed with Amodei, as did Elon Musk. Bill Gates has also expressed his reservations about the speed of development of artificial intelligence and the threat it poses to human society.

While society is beginning to wonder what a single robot could do to harm one or more humans, some are now asking a more fundamental question: Can human civilization be endangered by extinction due to artificial intelligence? Former Anthropic and OpenAI researcher Jacob Coxon said there was a 10% chance of human extinction occurring due to the rapid pace of agent AI development.

Evan Hubinger, head of anthropogenic alignment science, further increased the catastrophic odds. Asimov’s warning evidently didn’t go far enough.

When Asimov published his novel in 1985, Robots and Empirehe had concluded that his original three laws were incomplete. After years of reflection, he introduced what he called the Zeroth Law, which he deliberately called “Zero” because he believed it should come before the other three rather than simply become a new fourth law. Why? Because Asimov’s Zeroth Law expands the original prohibition against harming a single human being to encompass the well-being of humanity as a whole.

According to Asimov it therefore deserved to precede the First Law. Much of society’s concern about robots must now shift from the harm that a single robot might inflict on an individual to the potential harm that intelligent robots might inflict on civilization itself.

Even if AI does not endanger human civilization, the concern is very real that other socially harmful effects could result. The implications of the Zero Law remain unknown to the general population and are only now being mentioned by AI developers of the generation.

An example of this broader concern is the recent statement released by numerous economists and technologists titled “We must act now.” More than 2,500 prominent signatories have already endorsed the statement, calling on the company to take immediate action.

The signatories said that waiting much longer will not leave enough time to provide software developers with the guidance they need to moderate the otherwise disruptive ability of robots to replace human workers without allowing adequate time for transition and institution building.

We must act now if we are to provide sufficient guidance for the development of robotics so that human society as a whole is not negatively affected. We also need to start designing the transition tools needed to manage future changes.

In counterpoint, NVIDIA CEO Jensen Huang said that concerns about the dangers of artificial intelligence are overblown.

The proposed legislation follows Asimov’s lead

On July 21, 2026, Senator Mark Warner, Vice Chairman of the Senate Intelligence Committee, introduced comprehensive legislation titled “A Framework for the Future of America’s Artificial Intelligence,” consisting of three major components.

About two-thirds of Senator Warner’s proposal focuses on broad requirements that he says must be put in place now to avoid future harm. He is particularly concerned about the potential negative effects that robotics could have on human civilization if adequate controls are not established for the future development of artificial intelligence, as well as the need to prevent unauthorized commercial exploitation and protect the security of consumer data.

The second part of the legislation addresses the threat of labor market instability resulting from the rapid spread of AI systems across the economy without adequate planning or institutional support for the transition from the pre-AI to the post-AI world.

Human society could face a widespread economic crisis if it fails to establish adequate guardrails before robotics development advances further, including the provision of significant transition resources for the workforce. Titled the National Workforce Transition Fund, this piece of legislation would provide job retraining for workers, retention grants and education assistance for workers looking to move into new occupations as existing jobs are replaced by automation.

In explaining the need for this second part of his legislation, Senator Warner expressed his belief that within two years, approximately 30% of college graduates may not be able to find work after completing their studies due to the rapid expansion of artificial intelligence in the workplace. This estimate could turn out to be correct or, ultimately, excessively pessimistic. Only time will tell.

Warner’s broader warning, however, that we must begin developing programming safeguards and transition policies now, is well received. As he noted, if we wait another two years, it will probably be too late. This is exactly the point made in the statement “We must act now”.

Mid-level white-collar workers are the most vulnerable right now. It will likely take many years for robots to replace plumbers, carpenters, electricians and other skilled tradespeople. Unfortunately, many of us are already familiar with white-collar workers who have lost their jobs as AI systems have begun to replace human labor in the workplace.

The industrial revolution developed over a relatively long period of time compared to the extraordinary speed with which robotics and artificial intelligence are entering today’s society. Looking back, we can easily see how better planning of our social institutions during the Industrial Revolution would have made the transition to the postindustrial world easier for everyone and considerably more politically stable. Let’s not make the same mistake again.

David Sacks, President Trump’s artificial intelligence advisor, responded to industry-wide calls for a slowdown in the development of agentic AI by saying that the industry could agree among itself on guardrails and did not require administrative action. He would be right if the administration created an antitrust “safe harbor” exemption from Federal Trade Commission and Justice Department suits for such collusive conversations between competitors. The effect on global competition (e.g., China) of such a “slowdown” is a topic that needs to be addressed and understood quickly.

Ultimately, more jobs will likely be created in the post-robotic future than exist today, just as happened after the industrial revolution. To ensure that the transition is as orderly and humane as possible, following Asimov’s example, we must act now. Even more critically, if we want human workers to remain in control, we must act now.

About the author

Rinaldo Brutoco, founding president and CEO of the World Business Academy, is an entrepreneur, executive, author, radio host and futurist. He has published numerous articles and books addressing the role and responsibility of business in relation to the critical moral, environmental and social concerns of the day.

Brutoco is a regular conference keynote speaker and guest lecturer at business schools such as Stanford Business School, Columbia Graduate School of Business, Northwestern University’s Kellogg Graduate Business School, and North Carolina University’s Keenan-Flagler Graduate School of Business. For over 35 years, Rinaldo has been widely recognized as a practical visionary and change agent.