BEIJING, July 6 (Xinhua) -- Decision-makers and strategists attending the 13th World Peace Forum have sounded the alarm over the risks posed by artificial intelligence (AI).

The three-day event, which wrapped up on Friday in Beijing, put AI in the spotlight, featuring two dedicated panels that examined its risks and regulation, as well as its expanding role in armed conflict.

Regarding AI risk and regulatory challenges, a widely held view among participants was that the use of AI must never come at the expense of human safety, and that life-and-death decisions should always remain in human hands, rather than being delegated to machines.

"We have to be very careful that we don't unintentionally unleash the power of AI in a way that would harm people," said Balthasar Staehelin, personal envoy of the president to China for the International Committee of the Red Cross (ICRC), "there is an ethical responsibility here -- one that must always be considered in light of international humanitarian law."

Advancements in information technology have enabled AI's widespread use in conflict scenarios, yielding both promising and troubling results. The ICRC and other humanitarian organizations, for example, are deploying AI to match individuals searching for displaced family members, as their names may be spelled slightly differently in Arabic, and to help plan the best routes for refugees fleeing conflict zones.

AI has been used in mine clearance, too. Low-flying drones equipped with AI technology can help detect mines, a way far more efficient than previous methods.

Unfortunately, AI has also been employed to search for, track and identify combatants, and even lock onto targets, raising serious concerns about the growing role of robots in lethal operations, potentially even taking human lives.

This unease is hardly new. As early as 1942, science-fiction writer Isaac Asimov proposed his "Three Laws of Robotics," which included the rule: "A robot may not injure a human being or, through inaction, allow a human being to come to harm."

In an interview with Xinhua this March, Staehelin also noted that the act of killing in war should always be a human decision, not one made by AI.

This warning was echoed by Bruno Angelet, Belgian Ambassador to China, who believed that accountability for battlefield decisions must always rest with people, not machines. "It is a person who decides that the system can make decisions, and it is a person who is accountable. There is no stone or computer that will be accountable," he said.

Reinforcing this perspective, Zeng Yi, founding dean of Beijing Institute of AI Safety and Governance, observed that AI acts as a mirror -- reflecting humanity's choices and exposing whether our actions are truly responsible, especially when reckless decisions by individuals threaten to endanger society as a whole.

Xiao Qian, deputy head of the Center for International Security and Strategy at Tsinghua University, furthered the discussion by highlighting the legal ambiguity and difficulties in assigning accountability in AI-led actions.

If a drone misidentifies a target, it becomes extremely challenging to determine who is responsible -- the programmer, the commander, or the machine itself -- sparking ongoing debate. Many AI systems also operate as black boxes, making legal compliance and effective oversight nearly impossible, she said.

Christian Bessiere, research professor at the French National Center for Scientific Research and the University of Montpellier, broke down AI's risks into four main categories: economic, "black box" opacity, cultural, and military. On regulation, he cautioned against excessive oversight that could stifle innovation and advocated for the creation of a strong, independent, non-profit international body to provide regulatory guidance.

Turning to the question of global AI governance, Gong Ke, executive director of the Chinese Institute of New Generation Artificial Intelligence Development Strategies, identified escalating geopolitical tensions as the greatest risk facing international cooperation on AI.

He added that a key challenge is making AI governance more inclusive, ensuring that all countries and stakeholders have a seat at the table.

Yet even reaching the smallest consensus on AI governance remains extremely challenging amid ongoing geopolitical strains and competing national interests, observed Xiao Qian.

She noted that while scientific groups are actively launching initiatives to raise awareness about the risks of military AI, these efforts have yet to receive significant attention or respect from governments.

Themed "Advancing Global Peace and Prosperity: Shared Responsibility, Benefit and Achievement," the forum, which ran from July 2 to 4, was hosted by Tsinghua University and the Chinese People's Institute of Foreign Affairs.