Boox Adopts Censored AI Model Amid Shift in China

We independently review all recommendations. Purchases made through our links may earn us an affiliate commission. Here’s how it works.

Chinese E-Reader Boox Replaces GPT-3 with Censored AI Model Promoting Propaganda

Chinese e-reader manufacturer Boox has made a controversial shift by replacing Microsoft Azure’s GPT-3 in its devices with a language model developed by ByteDance, the parent company of TikTok. This decision has sparked widespread criticism due to the heavy censorship and state-aligned propaganda embedded in the new AI assistant.

Key Issues with Boox’s New AI Assistant

The new ByteDance-based AI model restricts discussions on sensitive issues, refuses to criticize China or its allies, and sidesteps questions that challenge state-approved narratives. Notable examples include:

  • Censorship of Sensitive Topics: Any mention of “Winnie the Pooh,” often used to satirize President Xi Jinping, is blocked.
  • Promoting State Narratives: When asked about Russia’s actions in Ukraine, the model frames it as a “complex geopolitical situation” influenced by NATO rather than addressing the core issues of the conflict.
  • Misinformation about Historical Events: On topics like Tiananmen Square, the AI avoids historical facts, substituting them with state-driven messaging.
  • Propaganda-Driven Responses: The model denies accusations such as the reported genocide of Uyghurs, aligning itself with Chinese government positions.

Such overt censorship raises concerns around the spread of state propaganda through technology exported to a global audience.

Users and Community Reactions

The switch has also frustrated users outside of China. Many reported that when they attempted to discuss these issues on Boox’s official Reddit forum, their posts were deleted. Furthermore, the AI assistant may currently be disabled, with reports suggesting it has been rendered unavailable. Boox has yet to issue an official statement clarifying its reasons or addressing user backlash.

Meanwhile, competition in the e-reader market is heating up, with alternative devices entering the fray. The upcoming Hanvon Clear 7 Turbo is positioning itself as a potential disruptor, while other premium e-readers like the Supernote A5 X2 Manta aim to challenge incumbents with advanced features emphasizing functionality and privacy.

Broader Implications of Imported AI Models

Boox’s adoption of a censored AI model highlights a growing issue: imported AI systems often carry the cultural, ethical, and political biases of their origin. As OpenAI CEO Sam Altman previously warned, companies leveraging foreign AI technology also “import” the values embedded within these systems.

In China, all AI models must pass strict government vetting to ensure they conform to “socialist values.” This alignment not only restricts domestic use but can also influence international adopters. Previous examples of such AI censorship include Baidu’s ERNIE-ViLG, which refuses to generate images related to Tiananmen Square, and other Chinese systems that misrepresent politically sensitive topics while generating permissible alternative content.

Consumers seeking alternative e-note devices might also take interest in non-traditional offerings such as the Guowen R8 E-Note, known for its stunning e-paper display and feature versatility, positioning it as a noteworthy contender for those concerned about balanced technology.

Growing Global Pushback

Some countries are taking proactive steps to counter such influences. Taiwan, for instance, is developing its own AI model, “Taide,” aimed at providing unbiased and reliable alternatives for businesses and governments.

The Boox controversy serves as a reminder that artificial intelligence carries more than just technological advancements; it brings the ideologies of those behind it. As AI becomes further embedded in everyday devices and platforms, global consumers must carefully scrutinize the origins and neutrality of the systems they adopt.

For readers exploring other e-reader options amidst this controversy, devices like the Kindle Paperwhite continue to offer reliable performance at competitive prices, appealing to users prioritizing a neutral reading experience over advancements like AI-driven assistance.

This case resonates as a critical moment in understanding how AI can be leveraged for ideological dissemination and the need for transparent, value-neutral AI deployment.

Leave a Reply

Your email address will not be published. Required fields are marked *

Back to top button