The Responsible AI Imperative

Enterprise-grade generative AI promises to revolutionize industries, and to harness the power of generative AI responsibly, grounding it in enterprise truth to foster Responsible AI is critical.

Grounding refers to the practice of connecting AI models to reliable information sources, ensuring that their outputs are based on verified data. It plays a pivotal role in making generative AI more accurate, reliable, and useful for a wide range of applications, and helps bridge the gap between the model's internal knowledge and the real world, leading to more informed and trustworthy AI systems.

For example, grounded chatbots can access company-specific knowledge bases or product documentation to provide accurate answers to customer inquiries. Companies using creative AI agents to develop marketing copy can also use grounded models and be assured that they are developing materials that are based on real-time market data or verified sources.

Enabling real time data access and adapting to dynamic environments

By grounding AI models in real-time information sources, such as news feeds or financial data, they can provide up-to-date and relevant responses to user queries. This is particularly valuable for applications like chatbots or virtual assistants.

Grounding also allows AI models to adapt to changing circumstances and information. For example, a grounded model used for navigation could take into account real-time traffic data to provide the most efficient route.

When customers select Grounding with Google Search for their Gemini model, Gemini will use Google Search, and generate an output that is grounded with the relevant search results. It is simple to use, and it makes the world’s knowledge available to Gemini. 

These capabilities address some of the most significant hurdles limiting the adoption of generative AI in the enterprise: the fact that models do not know information outside their training data, and the tendency of foundation models to “hallucinate,” or generate convincing yet factually inaccurate information. Retrieval Augmented Generation (RAG), a technique developed to mitigate these challenges, first “retrieves” facts about a question, then provides those facts to the model before it “generates” an answer – this is what we mean by grounding. Getting relevant facts quickly to augment a model's knowledge is ultimately a search problem.

Grounding with Google Search entails additional processing costs, but because Gemini’s training knowledge is very capable, grounding may not be needed for every query. To help customers balance the need for response quality with cost efficiency, Grounding with Google Search will soon offer dynamic retrieval, a novel capability that lets Gemini dynamically choose whether to ground user inquiries in Google Search or use the intrinsic knowledge of the models, which is more cost-efficient. 

The model does this based on its ability to understand which prompts are likely to be related to never-changing, slowly-changing, or fast-changing facts. Consider scenarios like inquiring about the latest movies, where Grounding with Google Search can provide the most up-to-date information. Conversely, for general questions, like "Tell me the capital of France,”, Gemini can instantly draw from its extensive knowledge, providing responses without the need for external grounding. 

Private data is not on the internet and Google Search wouldn’t be able to find it, so in addition to Grounding with Google Search, we offer multiple ways to apply Google-quality search to your enterprise data. Vertex AI Search works out-of-the-box for most enterprise use cases. And for customers looking to build custom RAG workflows, create semantic search engines, or simply upgrade existing search capabilities, we offer our search component APIs for RAG. This suite of APIs, now generally available, provides high-quality implementations for document parsing, embedding generation, semantic ranking, and grounded answer generation, as well as a fact checking service called check-grounding.

Supporting decision-making while improving reliability and trustworthiness

Grounded models are more likely to produce consistent and accurate responses, making them more reliable for critical applications where misinformation can have serious consequences. This increased reliability fosters trust in the technology.

The answers generated with RAG-based agents and apps typically merge the provided context from enterprise data with the model’s internal training. While this may be helpful for many use cases, like a travel assistant, industries like financial services, healthcare, and insurance often require the generated response to be sourced from only the provided context. Grounding with high-fidelity mode, now in experimental preview, is a new feature of the Grounded Generation API that is purpose-built to support such grounding use cases. 

The feature uses a Gemini 1.5 Flash model that has been fine-tuned to focus on customer-provided context to generate answers. The service supports key enterprise use cases such as summarization across multiple documents or data extraction against a corpus of financial data. This results in higher levels of factuality, and a reduction in hallucinations. When high-fidelity mode is enabled, sentences in the answer have sources attached to them, providing support for the stated claims. Grounding confidence scores are also provided.

To make it easier to use trusted third-party data for RAG, starting next quarter, Vertex AI will offer a new service that will let customers ground their models and AI agents with specialized third-party data from premier providers such as Moody’s, MSCI, Thomson Reuters, and Zoominfo. This will help enterprises integrate third-party data into their generative AI agents to unlock unique use cases, and drive greater enterprise truth across their AI experiences.

Reliable, robust, and trustworthy gen AI

In the rapidly evolving landscape of generative AI, grounding is the linchpin that separates hype from reality. Google Cloud's unwavering focus on grounding sets it apart as a leader in trustworthy and responsible AI, empowering enterprises to confidently embrace this transformative technology.

Annop Siritikul, Country Director, Thailand, Google Cloud