"Artificial intelligence has become a real danger to humanity, those in the field bear a duty of care," says Prof. Shmuel Itzkowitz, founder of the Faculty of Computer Science at the College of Management. According to him, a warning light was recently triggered from within the industry itself, with even OpenAI and Anthropic already calling to slow things down.
Prof. Itzkowitz explains that artificial intelligence is indeed changing the way human beings learn, work, and make decisions. It can improve medical diagnosis, make knowledge accessible, and streamline public services. However, processing data and predicting patterns do not guarantee a correct or fair decision. The question is not whether to stop development, but how to ensure technology remains a tool serving humanity, rather than a system determining human fate without accountability or appeal. The central risk is flawed decisions that mistakenly appear to us as objective.
"AI systems learn from historical data," he warns. "When that data includes discrimination, errors, or a lack of representation, they are liable to replicate distortions in decision-making regarding credit, employment, welfare, or the legal system. Compounding this is false information on a broad scale: Convincing texts, images, and voices enable impersonation, fraud, and influence over public discourse."
Prof. Itzkowitz adds that "the picture was sharpened over the past week when OpenAI announced that it supports binding requirements for independent safety testing, and that development should be slowed or stopped when its safety cannot be guaranteed. Anthropic reported the disruption of malicious uses of its models, from cyber and fraud to surveillance. The responsibility does not lie with engineers alone. Governments need to establish enforceable rules, educational institutions need to foster critical digital literacy, and Companies need to prioritize safety even if it delays rollout."