Microsoft Bing New AI New Controls With AI Training From AI Search Grounding

Microsoft Announces Bing AI Content Controls

Microsoft has introduced a set of refined content controls for its Bing AI chatbot. According to Krishna Madhavan, Principal Product Manager for Microsoft Bing, the new preferences are “newer, simpler, and clearer” and let users decide how their data is used—whether for training the AI, for grounding responses, or for both—while still remaining eligible for Bing search results.

What the Controls Do

• Training – Determines if user inputs can help improve future generations of the AI model.
• Grounding – Allows the system to reference real‑time search results to provide up‑to‑date answers.
• Both – Enables the model to learn from the interaction and also pull in live search data.

How to Manage Your Preferences

1. Open the Bing chat interface.

2. Click the gear icon or settings menu.

3. Toggle the training, grounding, or both options to your preference.

4. Save changes – the new settings apply immediately.

Transparency and Control

The updated controls give users a more intuitive way to understand how their data is handled. Rather than a single, opaque setting, the options are now broken down into distinct choices, helping users make informed decisions.

Why It Matters

For business owners, the ability to manage AI training data is critical. When customers interact with a company’s chatbot or search interface, the information they share can influence the AI’s future responses. By offering granular controls, Microsoft reduces the risk of unintended data exposure while still providing the benefits of AI‑enhanced search.

Moreover, the clarity of these settings aligns with growing regulatory expectations around data privacy. Companies that adopt Bing’s updated controls can demonstrate to stakeholders that they are actively managing user data in line with best practices.

Finally, the choice between training, grounding, or both gives organizations flexibility. A business that prioritizes real‑time accuracy might lean toward grounding, while one focused on continuous improvement could opt for training. This flexibility supports a range of use cases, from customer support to market research.

FAQs