Tencent launches new AI model with safety features
Istanbul (ANTARA) – Chinese technology company Tencent on Wednesday (4 November) launched a new artificial intelligence (AI) model called Hy4, equipped with safeguards designed to minimise the risk of misuse in cyberattacks and potentially dangerous biological research.
The Shenzhen-based company, widely known as the developer of the popular messaging app WeChat, said the model was trained on cybersecurity data focused on defence, such as identifying and fixing software vulnerabilities, without including material related to offensive exploitation.
Tencent said security assessments conducted by its internal research team found no significant concerns about the model’s ability to assist with cyberattacks in the areas tested.
The company also said that although Hy4 has a strong understanding of biology, its ability to design open-ended experiments is more limited than that of several other advanced AI models.
In addition to its safety features, Tencent said Hy4 can analyse images, interpret complex charts, write computer code and perform various tasks through software interfaces.
Tencent said the model is already available through several of its cloud services and workplace applications. The company also plans to release the model weights and technical documentation in the coming days.
Hy4 is the latest addition to Tencent’s growing AI portfolio, following the launch of a preview version last August.
According to the company, the latest model offers improved performance in several fields, including financial analysis, scientific research and software development.
Source: jawawa.id (https://jawawa.id/newsitem/tencent-launches-new-ai-model-with-safety-features-1793861661)
Originally published by Antara (Indonesian) (https://www.antaranews.com/berita/5833331/tencent-luncurkan-model-ai-baru-dengan-fitur-keamanan), translated from Indonesian