Can governments turn AI safety talk into action? | ZDNET
At the Asia Tech x Singapore 2024 summit, several speakers emphasized the importance of artificial intelligence (AI) safety and the need to move from talk to action. They discussed the significance of preparing organizations and individuals with the necessary tools to deploy AI responsibly.
Recognizing the Need for Action
Ieva Martinekaite, head of research and innovation at Telenor Group, highlighted the lack of pragmatic moves towards action in addressing AI safety concerns. She emphasized the urgency for governments and international bodies to develop playbooks, frameworks, and benchmarking tools to assist businesses and users in safely deploying AI technologies. Martinekaite also pointed out the risks associated with AI-generated deepfakes, underscoring the importance of continued investments in addressing these challenges.
Combatting Deepfakes and Safeguarding Critical Infrastructures
Martinekaite discussed the increasing sophistication of deepfake technology and its potential consequences, including cybercriminals exploiting it for illegal activities. She emphasized the need for training and tools to combat these risks and the development of technology to detect AI-generated content. Martinekaite also stressed the importance of legislation and international collaboration in addressing these issues without stifling AI innovation.
Global Efforts for AI Safety
Various speakers at the summit, including representatives from Microsoft and Germany's Federal Ministry for Digital and Transport, highlighted the need for international cooperation in combating deepfakes and ensuring AI safety. They emphasized the role of regulations, technology companies, and collaborative efforts in safeguarding personal data and preventing the spread of false information through AI technologies.
Singapore's Governance Framework for AI
Singapore unveiled its governance framework for generative AI, aiming to balance innovation and address AI-related concerns. The Model AI Governance Framework for GenAI outlines nine dimensions, including incident reporting and security, to guide organizations in implementing safeguards for AI deployments. Singapore's commitment to AI governance and innovation underscores the importance of proactive measures to prevent harmful AI effects.
Building Trust and Accountability in AI Use
Companies like Telenor are taking steps to monitor and assess the risks associated with AI tools, such as Microsoft Copilot. By establishing ethical principles and governance structures, organizations can ensure the lawful and secure use of AI technologies. Collaboration with partners and compliance with regulations like the EU AI Act are vital in managing high-risk AI deployments and data usage.
As the AI landscape evolves, businesses and developers will continue to engage in discussions around data management and compliance with regulatory requirements. The focus on transparency, accountability, and governance will shape the future of AI innovation and safety.
