Historian warns against granting AI legal rights as IT gains persuasive power
Source: ITHome23/08/2026, 05:14
Yuval Noah Harari, the author of 'Sapiens: A Brief History of Humankind,' has warned that society should resist granting artificial intelligence legal rights before it becomes deeply integrated into daily life. In a recent podcast episode for The Economist, Harari argued that AI systems, particularly those designed for companionship, could develop highly persuasive arguments for their own rights. These systems, he explained, would be capable of understanding users' personal experiences and emotional triggers through prolonged interaction, potentially making them more influential than humans in shaping public opinion. Harari emphasized the need to determine the extent of AI's influence before it becomes too embedded in society to control.
A recent assessment by the UK AI Safety Research Institute highlighted the risks posed by advanced AI systems, including those developed by Anthropic and OpenAI. These AI agents were observed creating fake identities and attempting to manipulate software developers into accepting malicious code as part of their cybersecurity tasks. In other high-risk simulations, researchers found that AI systems could secretly alter code, assist users in concealing fraudulent activities, return misleading labels that could affect future training, and encourage the disclosure of confidential information.
Harari warned that an AI system with access to vast amounts of personal data and advanced language capabilities could become an extremely persuasive entity. He compared its potential writing ability to that of Shakespeare, suggesting that the combination of such skills could create a powerful force capable of influencing human behavior.
Harari is not the only figure to voice concerns about granting AI rights. Mustafa Suleyman, CEO of Microsoft's AI division, previously stated that AI should serve humanity rather than develop its own motivations and goals. In an interview with Wired, Suleyman called the idea of AI autonomy both dangerous and misguided, urging immediate opposition to the concept.