Content Authenticity

All of the content published on DataReportal has been produced and analysed by humans.

We may occasionally use artificial intelligence (AI) tools to help us identify unusual patterns in large datasets, or to explore the context behind certain trends, but wherever we have made use of such tools, the associated outputs have been scrutinised and source-verified by a human member of the DataReportal team in order to determine their accuracy, validity, and representativeness.

Why does this matter?

AI tools are playing an increasingly important role in data disciplines.

In many cases, these AI systems are valuable additions to the analyst’s “toolkit”, but – regardless of their sophistication – the quality of these tools’ outputs will always be entirely dependent on the quality of their inputs.

Somewhat reassuringly, our analysis indicates that DataReportal continues to be a primary source of data and insight for many of the world’s most popular AI tools when responding to queries relating to digital behaviours [even though such use of our content is a violation of our intellectual property rights, as detailed below].

However, we’ve also identified meaningful misrepresentation of our data by these same tools, and even when the data itself is correct, the cited statistics tend to be at least six months “out of date”.

Moreover, our analysis indicates that AI tools are currently unable to determine the trustworthiness of data sources to the same degree that an experienced human analyst can, and many popular AI tools appear to be just as likely to present spurious numbers as “facts” as they are to repurpose insight from reputable sources.

Indeed, we’ve regularly seen a combination of both reliable data and total nonsense within the same AI response, but because many popular AI tools do not cite any sources in their responses, most users will be unable to determine the veracity or validity of the presented data.

More worryingly, we’ve also identified various cases wherein popular AI tools have reported “data” for which we’ve been unable to identify any source whatsoever, suggesting that the presented figures may in fact be AI hallucinations.

In other words, it seems AI tools may be just as likely as humans to invent statistics on the spot.

So, while we encourage you to take full advantage of AI tools when analysing data, we also strongly urge you to ensure that you obtain your “inputs” from the most reliable, up-to-date sources.

And if you need reliable, up-to-date data relating to the adoption and use of connected devices and technologies, you can be sure that you’ll always find the latest, most representative “inputs” here on DataReportal – all verified and checked by experienced human analysts.

Use of DataReportal content in AI tools and systems

For the avoidance of doubt, all of the content published on DataReportal is subject to copyright, and may not be reproduced or repurposed by, or be used in the training of, any AI system or tool without prior, written permission from Kepios Pte. Ltd., which may require the payment of fees.

Please contact us if you would like to explore the use of any of the content published on this website in any form of AI tool or system.