Why are Singapore firms hesitant to adopt DeepSeek despite cost savings?

Question

Accepted Answer

Singaporean Firms Exercise Caution with DeepSeek Amidst Data Security and Bias Concerns

Despite the significant cost-saving potential offered by DeepSeek, companies in Singapore are demonstrating a cautious approach to its adoption. This hesitation stems primarily from persistent concerns regarding data security and the potential for AI biases within the model. Major corporations, including financial institutions, consulting firms, and cybersecurity specialists, have implemented stringent policies that prohibit employees from utilising generative AI tools like DeepSeek for professional tasks without rigorous due diligence. This measured response underscores a broader industry trend towards prioritising data integrity and ethical AI deployment over immediate cost reductions.
Data Security and Confidentiality Risks Remain Paramount

A core reason for the apprehension among Singaporean businesses revolves around the inherent risks associated with data retention and potential corporate data leaks when using generative AI tools. While DeepSeek, as an open-source model, presents a compellingly low-cost alternative to competitors, its operational mechanisms have raised questions. Experts express concern about the extent to which the DeepSeek application stores and scrutinises user prompts and generated outputs, particularly for the purpose of further model training. This lack of transparency regarding data handling practices creates a significant hurdle for organisations that are legally and ethically bound to protect sensitive client information.
Concerns Over AI Biases and Responsible AI Standards

Beyond data security, the potential for AI biases within DeepSeek's outputs is another significant factor contributing to corporate hesitancy. Early assessments suggest that the model may not consistently adhere to established responsible AI standards. For instance, observations have been made regarding its ability to avoid answering questions on topics often subject to censorship in certain regions, while providing responses on similar issues in other contexts. This selective behaviour raises questions about the model's neutrality and its potential to perpetuate or embed biases, making its widespread adoption for critical business functions a point of concern for risk-averse organisations.
Regulatory Scrutiny and Precautionary Measures

The international regulatory landscape also plays a role in shaping Singaporean firms' cautious stance. Governments in South Korea and several European nations have formally requested information regarding DeepSeek's training data, indicating a heightened level of scrutiny. Furthermore, South Korea, Italy, and Australia have proactively blocked access to DeepSeek on government devices due to perceived security risks. Similarly, legal firms within Singapore have instituted prohibitions on the use of generative AI tools for handling client data until their safety and reliability can be thoroughly validated. This environment of regulatory caution encourages businesses to adopt a similarly precautionary approach, prioritising thorough risk assessment and compliance.

Reading history

Hide →

This history is cleared when the page is closed.

Why are Singapore firms hesitant to adopt DeepSeek despite cost savings?

Loading

Was this article helpful?

Helpful

Not helpful

Report an issue

1

0

0

Share with friends