Twitch adopts a stricter approach towards explicit deepfakes


Twitch already forbids explicit deepfake images and videos, but it's taking a tougher position against them today. The livestreaming service is updating its policy on adult nudity to include a ban on synthetic NCEI (non-consensual exploitative images), even if it's only shown briefly or to criticize its existence. It's also revising ts sexual violence and exploitation policies to make clear that intentionally making and sharing non-consensual deepfakes can lead to a ban with the first offense.

The policy changes should take effect within the next month. The company hopes the added clarity and modernized language will deter potential offenders. The company is holding a virtual Creator Camp with NCEI expert Zara Ward on March 14th to help spot and deal with malicious deepfakes. We've asked the company if it has any plans to detect deepfakes as they appear, and will let you know if we hear back.

The update comes after an incident in which well-known streamer Atrioc briefly showed a…

2023-03-07 12:54:07
Source from www.engadget.com
Twitch Adopts a Stricter Approach Towards Explicit Deepfakes

Artificial intelligence advancements have revolutionized various industries in recent years, including the entertainment industry. Deepfakes – manipulated videos that feature realistic images of individuals that did not actually occur – are particularly prominent in the world of streaming, with numerous streamers utilizing the technology to create compelling content. However, while deepfakes can be amusing, they also present ethical concerns and potentially result in harassment or defamation.

Twitch, a live-streaming platform for gamers, recognized these concerns and has opted to adopt a stricter approach towards explicit deepfakes in its user-generated content. In simplified terms, deepfakes can be created by training a computer model on a substantial quantity of videos and images of an individual and then using the information to produce new footage with the person’s face and speech. The severe consequences arise when creators use these deepfakes to defame, outrage, or harass others.

Twitch has become increasingly concerned that the deepfakes could pose a significant risk to the security and privacy of its users. As such, it has adopted stricter rules that push back against the use of explicit deepfakes in live streams. The new guidelines will require creators to obtain explicit consent from the individual they desire to use in a deepfake before producing any content. Additionally, Twitch prohibited the creation and sharing of malicious deepfakes, such as those intended to harm, embarrass or intimidate viewers of the platform.

While these new guidelines will help to address ethical issues and safeguard user privacy, Twitch concedes that enforcing them is complex. To combat this, Twitch will rely on both automated systems and human moderators to identify and intervene where necessary. The platform has already started to roll out more robust automated systems that can recognize deepfakes, as well as methods to escalate scenarios to human moderators.

Moreover, Twitch has also urged content creators to act responsibly and avoid using deepfakes in ways that can harm others. The platform further invited its users to report any deepfake-related violations promptly, which would undoubtedly enable swift action and help Twitch to enforce its new strict policies. With these policy changes and tighter surveillance, Twitch aims to protect its user base from the potential harm of deepfakes and ensure that its platform is safe and secure for viewers and broadcasters alike.

In conclusion, Twitch approaching deepfakes is both commendable and necessary, given the potential risks posed to users. As a platform that encourages creators to express themselves, it must also ensure that any content, particularly that which may cause harm, is policed effectively. Given the difficulties in enforcing such regulations, Twitch’s automated systems and human moderators will be keeping a close eye on content, so toxicity and explicit deepfakes can be extinguished promptly in favor of responsible and engaging content that upholds the community’s health and safety.

Exit mobile version