Build 2023 is finally here. And as expected, this year’s announcements mainly revolve around generative AI. While kick-starting this year, Microsoft debuted the new Bing, its AI copilot for the web to help enhance the browsing experience for users.
The company has since gone on to make a host of announcements highlighting the injection of AI capabilities across its products and services. In March, Microsoft released Bing Image Creator in creator mode, a tool designed to help users create images using their own words.
It’s in place to enhance the new Bing and Edge preview experience by providing users with better search, complete answers, a new chat experience, and the ability to create content. The Bing Image Creator that’s powered by OpenAI’s DALL∙E model lets users create images using their own words.
The thing with images generated using Bing Image Creator is that it’s very easy for a user to create images and use them elsewhere without attribution. However, Microsoft is now putting elaborate measures in place that will let users identify images and videos generate using AI both in Microsoft Designer and Bing Image Creator in accordance with the C2PA standard.
For those not conversant with Microsoft Designer, it’s an artificially intelligent social media content platform that can be sued to generate online content such as social media posts, greeting cards, branding, alerts, promotions, and more. It was first teased back in 2022, and shipped to preview earlier this year in January.
The new media provenance capabilities leverage cryptographic methods to watermark AI-generated content with metadata indicating its original source. At OnMSFT.com, we use Bing Image Creator to generate some of our featured images. Up until now, we’ve been using the tool to generate images and crediting the source at the bottom of the post.
These new capabilities will completely change how we interact with the tool. Microsoft has indicated that it is rolling out image and video formats support for Microsoft Designer and Bing Image Creator. It further added that these capabilities are expected to ship for users in the next couple of months.
Azure AI Content Safety & prompt flow
Additionally, Microsoft has also affirmed its commitment to developing AI technology responsibly, thus allowing users are able to make the most out of these new capabilities while simultaneously ensuring that they maintain responsible practices.
To further build on this premise, Microsoft has announced several new updates including Azure AI Content Safety, a tool designed to help users maintain and create a safe online environment. As such, the company has also indicated that it will be integrating the tool across Microsoft products
What’s more, the company has also announced new tools to Azure Machine Learning which include Responsible AI dashboard support for text and image data in preview.
The tools will enable users to better evaluate large models built using unstructured data during the model building and training phase. This way, it will be easier for users to spot errors as well as model explanations before proceeding to the deployment stage.
Also shipping to preview soon, prompt flow. A tool in place to help provide users with “a streamlined experience for prompting, evaluating, and tuning large language models.” Furthermore, the tool will also be integrated into the Azure AI content Safety tool, thus allowing users to easily detect and scrap harmful content directly within the flow of their work.