廣告
xx
xx
"
"
回到網頁上方
tvbs logo

Amid AI warfare surge, human oversight remains key: Experts

Reporter Dimitri Bruyas
Release time:2024/11/10 13:15
Last update time:2024/11/15 16:52
  • S

  • M

  • L

TAIPEI (TVBS News) — The rapid evolution of artificial intelligence (AI) in warfare is reshaping the landscape of physical and cyber conflicts. At the heart of this transformation lies a pressing concern: Can human cognition keep up with AI's capabilities? This question was a focal point at the KAS Australia and the Pacific Security Conference 2024 in Canberra, Australia, on Oct. 28, where experts delved into the consequences of AI's expanding role in global security operations.

AI and the Cybersecurity Dilemma

 
Professor Shiri Krebs, director of the Centre for Law as Protection at Deakin University, highlighted the clandestine nature of the AI arms race. "One thing that I mentioned in my comments is my concern from a cybersecurity dilemma," Krebs said, emphasizing the challenge of gauging other nations' capabilities. Unlike nuclear weapons, which are often tested openly, AI technologies remain veiled, fostering an atmosphere of assumption and uncertainty.

Peter Anstee, first assistant secretary of Australia's Cyber and Technology Security Policy Division, echoed the necessity of adopting AI tools in both warfare and intelligence contexts. "There's a strong appetite and necessity to adopt these tools as well," Anstee noted. However, he stressed the importance of retaining human decision-makers within the operational chain, as fully autonomous systems in cyber operations remain a distant reality.

The Human Element in AI Warfare
Krebs further explored the complexity of integrating human oversight in AI-driven warfare, posing critical questions about the human ability to manage AI systems. "A company that produces drones for both the two conflicts, in Gaza and Ukraine... advertises kind of swarm drones," Krebs explained, raising concerns about the cognitive demands placed on humans tasked with overseeing vast numbers of autonomous drones. As AI technologies continue to multiply threats with unprecedented speed, the need for effective strategies becomes paramount.
 

Helge Janicke, deputy CEO of the Cybersecurity Cooperative Research Centre at Edith Cowan University in Perth, emphasized the primary focus of organizations under cyber attack. "At the point where you're hit by an attack, the attribution is not the primary objective," Janicke stated. Instead, rapid response and containment are crucial to mitigating the impact on customers and operations, highlighting the essential role of speed and strategy in countering cyber threats in the AI era.

Janicke further highlighted the evolving nature of cyber threats, particularly the shift from ransomware to data exfiltration. "A cyber attack is an attack that an adversary is using digital means to break into an organization and extract some benefit from it," Janicke explained, underscoring the need for robust cybersecurity measures.

Legislation and Ethical AI Use
As AI's role in warfare continues to grow, the urgency for comprehensive legislation becomes increasingly apparent. The widening gap between AI capabilities and human cognitive limits demands robust regulatory frameworks to ensure the ethical and effective use of these powerful tools.

The discussions at the Konrad-Adenauer-Stiftung (KAS) conference underscored the necessity for nations to invest in AI technologies and develop policies ensuring the ethical and effective use of AI in military applications. Integrating AI into warfare extends beyond simple yes or no answers, requiring a nuanced understanding of the balance between technological advancement and human intervention.

To address these challenges, Janicke and his team have developed a board game, "Corporate Compromised," which educates players on effective cybersecurity strategies. This simulation guides participants through realistic scenarios, such as ransomware attacks, to demonstrate how adversaries exploit digital vulnerabilities for extortion.

The future of warfare may be uncertain, but thoughtful preparation and legislation could be our best defense against AI's evolving threats. As Krebs aptly noted, the question is not merely about having a human in the loop but understanding what that human can achieve in the face of AI's rapid advancements.