The authorities in the PRC have completed work on a document regulating the use of recommendation algorithms by tech companies. The rules will take effect on 1 March 2022, CNBC reports.
According to some provisions of China’s algorithmic regulation:
- Recommendation algorithms must not violate PRC law, for example by threatening national security.
- News services must obtain a license and may not disseminate fake news;
- Companies must inform users about the ‘principles, purpose and main mechanism of operation’ of the recommendation algorithms;
- Users must have the option to opt out of receiving recommendations generated by algorithms;
- Users must be able to select or remove the tags that are used to enable the recommendation algorithms and offer them something;
- Companies must promote ‘safe use’ of recommendations for the elderly, protecting them from fraud.
Violations carry fines of 10,000 to 100,000 yuan ($1,570–$15,740).
Kendra Shefer, a Beijing partner at Trivium China, says authorities could request access to the recommendation algorithms to verify their compliance.
“Algorithms are the companies’ most closely guarded secret, their most valuable asset, and letting the government rummage around in them would be problematic,” she said.
According to Shefer, the changes reflect some of the most pressing challenges facing modern Chinese society, among them internet control, the transparency of tech giants and anti-competitive behaviour.
As reported in August 2021, the Cyberspace Administration of China published rules for the use of recommendation algorithms by tech companies.
In October, the PRC’s Ministry of Science and Technology introduced a set of ethical principles for regulating artificial intelligence.
Subscribe to ForkLog’s Telegram news: ForkLog AI — all the news from the world of AI!