Artificial intelligence (AI) capabilities are expanding, but a new United Nations (UN) report says AI safety rules are lacking.
A preliminary report by the United Nations Independent International Scientific Panel on Artificial Intelligence, released on Wednesday (July 1), said time is running out to develop effective global governance of AI.
“This technology is transformative, but if the world continues on this trajectory, humanity will not realize the promised benefits,” Maria Ressa, co-chair of the Independent International Scientific Commission on Artificial Intelligence, said in a news release accompanying the report. “The risks to society, security, and our species are too high. The forces that drive AI forward are not the forces that deliver its benefits.”
The report points to several AI-related successes, including the use of technology to identify food insecurity and breakthrough advances in medicine and healthcare.
“AI has predicted the structures of more than 200 million proteins, accelerating drug discovery, vaccine development, and antibiotic resistance research,” the report said. “Doctors are using AI to detect diseases like breast cancer early, and healthcare workers in developing countries are leveraging AI tools in local languages to improve patient care.”
At the same time, the report says AI is allowing the spread of sexually abusive content and sexually explicit deepfakes, putting women and children at greatest risk.
The UN also pointed to the role of AI in spreading disinformation and exacerbating mental health problems, and its use by criminals “to carry out cyber-attacks, fraud and social engineering scams.”
The gap between rapidly advancing AI capabilities and effective risk management methods could someday have “catastrophic consequences,” the report said.
“There is a lack of reliable methods to maintain control of highly autonomous AI systems,” the report states. “There is no scientific guarantee that AI agents will not violate instructions, and there is already accumulating evidence that they do.”
To illustrate the speed of AI adoption compared to other technologies, the report noted that while it took decades for electricity to reach most homes, the internet reached the 1 billion user mark 15 years later.
“ChatGPT reached 100 million users in two months,” the report states. “Traditional policymaking has not caught up.”
