The UK government possesses vital datasets that could enhance AI applications, yet recent research by the Open Data Institute (ODI) unveils critical flaws in how these datasets are managed. Foundation models such as ChatGPT often access government data, but their outputs are frequently inaccurate due to reliance on secondary sources. This situation poses risks for citizens seeking guidance from AI on public services. The AI Opportunities Action Plan anticipates improvements via the National Data Library, but significant hurdles in dataset quality must be overcome to realize these benefits.
The UK's government datasets hold immense potential for AI innovations, but recent findings reveal that many AI models fail to accurately process this information.
The ODI's research shows that while foundation models use government data, they often rely on unreliable sources, compromising the integrity of AI outputs.
Collection
[
|
...
]