Tech Giant Allows Users to Create Deepfakes in Just 3 Minutes

Leading Chinese technology company Tencent has launched a platform that allows users to create deepfakes based on a three-minute video clip and 100-sentence voice material for 1,000 yuan ($145). The platform was developed by its subsidiary, Tencent Cloud.

According to a local media report, the deepfake platform relies on Tencent’s in-house artificial intelligence (AI) capabilities and technical abilities. The report noted that the platform could produce a convincing deepfake video in 24 hours in Chinese or English.

Tencent’s Platform Targeted at Informercials

Meanwhile, The Register reported that digital humans come in several styles, including 3D realistic, 3D cartoon, 3D semirealistic, 2D cartoon, and 2D real person.

Tencent plans to use the platform to host live-streamed infomercials for Chinese audiences. Thus, it will support displaying full and half-body images and appropriate gestures and movements.

Additionally, users can change the recording background, making it suitable for use in several commercial scenarios, including live streaming.

The company also provides a broadcasting platform where deepfakes can be operated and broadcast. The platform “supports the rapid generation of digital wisdom video through text and voice input,” Jiemian reported. 

Tencent Cloud executive in charge of the platform Chen Lei said that the company hopes to build an artificial intelligence-powered digital Homo sapiens factory.

The development marks a major milestone for Chinese tech companies trying to commercialize AI. Baidu has already released its Chatbot, while companies like Alibaba and Huawei are developing their AI versions.

Deepfake Poses Dangers

With AI’s increasing popularity and usage, a platform that allows users to easily create deepfake videos raises concerns about the potential for abuse. Earlier in the year, a US-based firm Graphika reported that China was using deepfake news anchors to spread political propaganda.

A separate Wall Street Journal report noted that the number of deepfakes online is almost certainly in the millions, highlighting the dangers this tool poses to the broader society.

Meanwhile, several stakeholders have already complained about how deepfakes are being used for scams and the spread of misinformation. This means that large-scale accessibility would likely force regulators to take more drastic actions to regulate AI usage.