The global artificial intelligence (AI) arms race is heating up, but AI’s development poses risks that governments around the world are seriously concerned about.
The European Parliament recently approved the EU AI Act — the first-ever legal framework on AI — which aims to address the risks of AI, including discrimination, privacy invasion and other potential problems.
AI emotion recognition would be prohibited in workplaces and education institutions, the EU legislation said. High-risk AI practices, such as utilizing AI to filter job applicants, would also be forbidden.
Some argue that applying overly stringent regulations on AI puts obstacles in the way of its development.
However, AI has the potential to cause severe problems just because of one single mistake. A car crash due to an error in a self-driving car is an example.
Underlying moral and ethical issues such as algorithm bias or criminal exploitation of AI technologies remain unresolved.
Deepfakes were originally designed to help complete the signal of an image, but criminals have co-opted the technology for more nefarious uses, such as the creation of fake celebrity porn.
Taiwan is the country most impacted by the dissemination of false information by foreign sources and has topped the Foreign Governments Dissemination of False Information indicator for a decade. With the increase in AI deepfakes, fake videos are everywhere, influencing public opinion and increasing cybersecurity concerns.
This being the case, how are we to ensure a safe AI industry in Taiwan? After all, we know it is easy for people with bad intentions to use AI to achieve their objectives. That poses a threat to our national security.
AI is a double-edged sword. It benefits people, but also has a negative impact on the human world. That is why governments and global institutions are taking steps to regulate AI.
In Australia, the government announced that it would monitor AI to prevent improper content generated by it. In the UK, the government has established the AI Safety Institute, the world’s first AI safety organization.
The EU AI Act, on the other hand is more comprehensive and practical than any other similar existing law or regulations. It would be a stepping stone for Western countries to build a legal framework for AI since the US still has implemented little AI legislation. It is also believed to be effective for countries and industries inside and outside the EU.
Taiwan plays a crucial role in the global information and communications technology supply chain.
It is time for our government to learn from other countries and consider how to legislate AI so that we can ensure the safe and sustainable development of the AI industry in Taiwan.
Dino Wei works in the information technology industry.
Translated by Hsieh Yi-ching
RSS