Microsoft recently announced an AI feature called "Correction," which detects and corrects AI hallucinations. Hallucinations are cases where AI models give out false or misleading information.
This feature is being added to the system for detecting the AI’s groundedness. AI’s groudedness is nothing but its ability to be in touch with reality. Azure AI Content Safety will have this feature to make sure AI gives accurate answers.
The new Microsoft feature, Correction, is specifically engineered to identify incorrect or unreal statements produced by AI systems in real-time. The feature as its name describes will correct such wrongful information to enhance user experience.
Correction Besides grounding documents would be a resource base that AI models use as a repository. This would help ensure the output matches factual data. Whenever the AI system generates a response, the grounding system checks whether the information produced is correct or contains errors. Correction will thus trigger a review if they do detect any misinformation, providing rightful suggestions.
Moreover, the user would get control over choosing a reasoning explanation that describes why the AI system flagged specific content as wrong. This would help make the correction process more transparent.
Correction is currently available only on Azure but is particularly suited to enterprise clients. This is owing to its necessity for the grounding documents that it uses in the summarization of long documents and complex question-answering scenarios. It's specifically useful in RAG-based model environments.
According to a Microsoft spokesperson, the feature uses both small and large language models to align output from generative AI. Of course, it can't guarantee perfection, but the system severely minimizes the possibility of companies utilizing unverified or otherwise wrong AI-generated content.
Other upgrades that the company is working on will include better-robust AI systems, as well as better content moderation tools. The upgrades will be rolled out over the next few months.
Correction feature is the company’s wider effort to enhance its user experience and content safety. Microsoft released video-based user verification to protect customers' credentials against phishing attacks, to make more than 95% of its customers safer.