\n<\/p>\n
One of the more unexpected products to launch out of the Microsoft Ignite 2023 event<\/a>\u00a0is a tool that can create a photorealistic avatar of a person and animate that avatar saying things that the person didn\u2019t necessarily say.<\/p>\n Called Azure AI Speech text to speech avatar, the new feature, available in public preview as of today, lets users generate videos of an avatar speaking by uploading images of a person they wish the avatar to resemble and writing a script. Microsoft\u2019s tool trains a model to drive the animation, while a separate text-to-speech model \u2014 either prebuilt or trained on the person\u2019s voice \u2014 \u201creads\u201d the script aloud.<\/p>\n \u201cWith text to speech avatar, users can more efficiently create video \u2026 to build training videos, product introductions, customer testimonials [and so on] simply with text input,\u201d writes Microsoft in a blog post<\/a>. \u201cYou can use the avatar to build conversational agents, virtual assistants, chatbots and more.\u201d<\/p>\n Avatars can speak in multiple languages. And, for chatbot scenarios, they can tap AI models like OpenAI\u2019s GPT-3.5 to respond to off-script questions from customers.<\/p>\n Now, there are countless ways such a tool could be abused \u2014 which Microsoft to its credit realizes. (Similar avatar-generating tech from AI startup Synthesia has been misused<\/a>\u00a0to produce propaganda in Venezuela and\u00a0false<\/a> news reports promoted by pro-China social media accounts.) Most Azure subscribers will only be able to access prebuilt \u2014 not custom \u2014 avatars at launch; custom avatars are currently a \u201climited access\u201d capability available by registration only and \u201conly for certain use cases,\u201d Microsoft says.<\/p>\n But the feature\u00a0raises a host of uncomfortable ethical questions.<\/p>\n One of the major sticking points in the recent SAG-AFTRA strike was the use of AI to create digital likenesses. Studios ultimately agreed to pay actors for their AI-generated likenesses. But what about Microsoft and its customers?<\/p>\n I asked Microsoft its position on companies using actors\u2019 likenesses without, in the actors\u2019 views, proper compensation or even notification. The company didn\u2019t respond \u2014 nor did it say whether it would require that companies label avatars as AI-generated, like YouTube<\/a> and a growing number<\/a> of other platforms.<\/p>\n Microsoft appears to have more guardrails around a related generative AI tool, personal voice, that\u2019s also launching at Ignite.<\/p>\n Personal voice, a new capability within Microsoft\u2019s custom neural voice service, can replicate a user\u2019s voice in a few seconds provided a one-minute speech sample as an audio prompt. Microsoft pitches it as a way to create personalized voice assistants, dub content into different languages and generate bespoke narrations for stories, audio books and podcasts.<\/p>\n To ward off potential legal headaches, Microsoft\u2019s requiring that users give \u201cexplicit consent\u201d in the form of a recorded statement before a customer can use personal voice to synthesize their voices. Access to the feature is gated behind a registration form for the time being, and customers must agree to use personal voice only in applications \u201cwhere the voice does not read user-generated or open-ended content.\u201d<\/p>\n \u201cVoice model usage must remain within an application and output must not be publishable or shareable from the application,\u201d Microsoft writes in a blog post. \u201c[C]ustomers who meet limited access eligibility criteria maintain sole control over the creation of, access to and use of the voice models and their output [where it concerns] dubbing for films, TV, video and audio for entertainment scenarios only.\u201d<\/p>\n Microsoft didn\u2019t answer TechCrunch\u2019s questions about how actors might be compensated for their personal voice contributions \u2014 or whether it plans to implement any sort of watermarking tech so that AI-generated voices might be more easily identified.<\/p>\n For more Microsoft Ignite 2023 coverage:<\/p>\nPersonal voice<\/h2>\n