Vertex AI expands grounding to reduce AI hallucinations
Google Cloud is expanding grounding capabilities in Vertex AI, enhancing response accuracy for generative AI models. This includes broader access to Google Search and new methods for managing "hallucinations".
What happened?
Google Cloud has unveiled expanded "grounding" capabilities within its Vertex AI platform. These updates aim to improve the ability of generative AI models to deliver fact-based responses. Among the updates is the general availability of grounding with Google Search, complemented by dynamic retrieval that balances quality with cost-effectiveness.
Key facts
”We are expanding these grounding capabilities to help our customers build more capable agents and apps.”
Why it matters
These enhancements are critical for building reliable AI applications. By grounding AI model responses in external, trusted data sources, the risk of "hallucinations"—where AI generates incorrect or non-existent information—is reduced. This benefits enterprises seeking to implement AI with high requirements for factual accuracy.
Who is affected?
Developers and enterprises using Google Cloud's Vertex AI platform are directly affected, gaining access to more advanced tools for creating robust AI agents and applications. End-users of these AI applications will experience more reliable and fact-based interactions. Companies in finance and law utilising third-party data such as Moody's or Thomson Reuters will particularly benefit from the new integration possibilities.
Impact on the EU
Grounding with Google Search is now generally available in Vertex AI globally, including the EU market. Future third-party data integrations may be subject to EU data legislation depending on the nature of the data, but the core grounding features are currently available.
What else you should know
Furthermore, a "high-fidelity" mode for experimental grounding is being introduced to further reduce hallucinations. Support for grounding with third-party data, including Moody's and Thomson Reuters, is expected in the third quarter of this year. Vector Search is also being expanded with hybrid search, now in public preview.
Quick answers about this story
Vad har hänt?
När hände det?
Varför spelar det roll?
The link opens in a new window and leads to the publisher's own site.
Källan har spårats automatiskt från utgivaren via Aheadlines signalkedja.
AI-verktyg i artikeln
Topics
Get similar news straight to your inbox
The reader's room
Send in a question or an addition. The newsroom reads everything before it's published and replies when relevant. No AI-generated text – just people.
Sign in to submit a comment or question.
Read the article through your role
- Assess technical risk: model choice, vendor lock-in, data flow and running cost.
- Update the architecture doc if new APIs or regulations touch production.
- Ensure observability + rollback plan before rolling out to production.
Generated angle — not editorial analysis of "Vertex AI expands grounding to reduce AI hallucinations"