Tencent Launches Hy4 AI With Built-In Cyberattack, Bio Safeguards
News originally published by
MENAFNThe Shenzhen-based company, best known as the developer of the widely used messaging app WeChat, said Hy4 was trained on cybersecurity data centered on defensive work, such as finding and patching software vulnerabilities. Material tied to offensive exploitation was left out.
Internal research teams at Tencent carried out security assessments and found no significant concerns about the model’s capacity to assist in cyberattacks across the areas tested, the company said.
On biology, Tencent said Hy4 showed strong knowledge of the subject but weaker abilities in designing open-ended experiments than some of the most advanced AI models.
Beyond its safety measures, Hy4 can analyze images, interpret complex charts, write computer code and carry out tasks through software interfaces, according to Tencent.
The model is already accessible through several of the company’s cloud and workplace applications. Tencent plans to publish its underlying model weights and technical documentation in the coming days.
Hy4 is the newest entry in Tencent’s expanding AI lineup, following a preview version released in August. The company said the new model delivers better performance in financial analysis, scientific research and software development.
Legal Disclaimer:
MENAFN provides the
information “as is” without warranty of any kind. We do not accept
any responsibility or liability for the accuracy, content, images,
videos, licenses, completeness, legality, or reliability of the information
contained in this article. If you have any complaints or copyright
issues related to this article, kindly contact the provider above.
Legal Disclaimer:
EIN Presswire provides this news content "as is" without warranty of any kind. We do not accept any responsibility or liability for the accuracy, content, images, videos, licenses, completeness, legality, or reliability of the information contained in this article. If you have any complaints or copyright issues related to this article, kindly contact the author above.
