An Israeli diplomat delivering greetings in eight languages to audiences around the world in their native tongues is just the beginning of how artificial intelligence can be used as a tool to augment the venerable practice of diplomacy.
Earlier this year, Ambassador David Saranga, head of the digital diplomacy division at the Foreign Ministry, posted videos on Twitter in which his likeness delivered a message in Arabic, Chinese Mandarin, Persian, Greek, Hindi, Portuguese, Russian and Turkish. Saranga does not speak these languages but wanted to encourage dialogue.
“I always wanted to communicate with you in the Turkish language, a language I heard in my childhood at my father’s house,” Saranga’s likeness said in the video addressing Turkey. “Now this is possible thanks to Israeli artificial intelligence technology, which allows me to speak in any language.”
Saranga was using a generative AI video program developed by Tel Aviv-based D-ID that can generate digital human-like multi-language-speaking avatars, making Israel one of the first countries to utilize AI in its digital diplomacy activities, according to the ministry.
“We understand that there is a revolution right now, that the entire digital sphere is changing,” Saranga told The Times of Israel. “And therefore, we are encouraging our people – and when I say our people I mean the embassies and the diplomats and so on and so forth – to dive into the AI world in order to see how we can use it for the public.”
The Foreign Ministry opened its first Twitter account in 2009. Today, the ministry maintains social media accounts in more than 50 languages, reaching around 2 billion impressions a year and providing Israel with significant global reach during diplomatically precarious periods.
Saranga noted that the Persian-language content created by the ministry on social media generated nearly 450 million engagements in 2022, with 93% of those viewers from Iran itself at a time when mass protests following the death of Mahsa Amini were roiling the Islamic Republic.
However, Saranga said, the use of the metaverse — a virtual reality in which users can interact with each other and with the environment — and AI in digital diplomacy is still in its nascent stage. In 2022, the Israeli Embassy in South Korea opened a diplomatic mission in the metaverse, which it claimed was the first of its kind.
“These two fields of metaverse and AI, we are still at the beginning of it,” Saranga said. “We are very advanced when it comes to delivering the Israeli message.”
هوش مصنوعی دیپلماسی:
به لطف فنآوری توسعه داده شده توسط شرکت اسرائیلی @D_ID_ اکنون میتوانیم با مخاطبان در سراسر جهان به زبان مادری آنها ارتباط برقرار کنیم!
— اسرائیل به فارسی (@IsraelPersian) January 24, 2023
To bypass the strict social media bans imposed by Iran, the ministry has also turned to Instagram.
“There are many small businesses in Iran who are doing business through Instagram, this is why the regime didn’t want to shut it down completely. So Instagram has become the most important platform to convey messages,” Saranga said.
The ministry claims it was among the first in the world to join TikTok in 2022, where demographics tend to skew young. “When it comes to public diplomacy, you find the public as soon as possible or you approach the public as soon as possible, which means even when they are teenagers, even before they set up their minds. And this was our thought at the very beginning, that we want to approach also the younger generation, because they consume news on TikTok.”
Tapping into Israeli tech
Using D-ID’s platform back in January, the Foreign Ministry thanked the Israeli startup for giving it the ability to converse with audiences around the world in their native languages, pushing the boundaries of traditional diplomacy.
“We know everyone’s talking about #ChatGPT but we have officially found our next AI obsession,” the ministry tweeted.
Saranga’s Twitter videos, with their multilingual messages appealing directly to key allies in the diplomatic arena, were produced using D-ID’s speaking portrait platform. The startup, which was founded in 2017, has found itself in the midst of several viral projects at the epicenter of the generative AI field by utilizing text-to-video generation of digital presenters or avatars, and incorporating text and image generation.
“D-ID is a key building block in the global generative AI space,” said D-ID co-founder and CEO Gil Perry. The technology “is a major gateway for new users to discover and experiment with AI tools in general and AI videos in particular.”
D-ID is well aware of the pitfalls that can bedevil such powerful technology, and says that it abides by ethical regulations set up by the Partnership on AI and the Content Authenticity Initiative.
“To prevent the spread of misinformation, we utilize detection algorithms and specialty third-party software,” Perry said. “We add digital watermarks to all the videos produced on our platform to clarify that the image is computer-generated.”
Yaniv Levi, VP of product marketing at D-ID, told The Times of Israel that there are limits placed on the type of content that can be uploaded to the speaking portrait technology, including bans on nudity, profanity, and the use of famous politicians’ likenesses.
Additionally, D-ID said that it is “dedicated to using its powerful technology for ethical, positive purposes,” including social impact campaigns, such as raising awareness about issues like domestic violence, HIV, and Holocaust education.
D-ID is also making strides in the metaverse, where users can create avatars to interact in the virtual world.
“We are metaverse-ready. Our products are already being used there, primarily for educational purposes,” Perry said. “Looking ahead, we aim to constantly improve the quality and control users have over the digital people they create.”
This month, D-ID unveiled a platform that allows users to simulate face-to-face conversation with AI-generated digital humans.