Hi,
I’m researching SyRI this week. No no, not this Siri:
SyRI stands for System Risk Indication, which was legally established in the Netherlands in 2014. Its purpose: to detect fraud with social benefits, taxes and other government provisions.
SyRI combines data from all kinds of agencies, such as the Tax and Customs Administration, the Social Insurance Bank and the UWV (the Employee Insurance Agency). This data is fed into a large algorithm tombola, which spits out high-risk individuals.
Despite critical reactions from both the Dutch Data Protection Authority and the Council of State, it easily passed through parliament.
Nevertheless, SyRI could soon be a thing of the past, as opponents took the state to court. The verdict comes next Wednesday 5 February.
What’s the problem?
First of all, if you’re not Dutch, why should you care? Internationally, this case is one of the first cases of its kind, so it might set a precedent in the Netherlands and abroad.
Even United Nations special rapporteur Philip Alston interfered in the debate last year. In a brief, he writes that SyRI poses "significant potential threats to human rights, in particular for the poorest in society".
So what’s the problem with SyRI? The plaintiffs claim you are "suspicious in advance".
Data collected about you by another organisation and for a different purpose – for example, your tax return – can now suddenly be used to find out whether you are fishy without you knowing it and without it being clear how the algorithm works ... because that’s secret.
Until now, the system has only been used in poor neighbourhoods, which only increases the chance of discrimination. And does it even work? Mmm. A Dutch newspaper revealed that so far no fraud has been detected with SyRI.
We’ll know more next week. Stay tuned.
#NerdAlert
There is now a Google for datasets. On Dataset Search, you can Google data from academic articles, from institutions such as the World Bank, and anyone else who puts data online in a format that’s readable for Google.
This makes my data heart beat faster. What patterns can be found in these 250,000+ lyrics? Or in this data from, in my opinion, the best Christmas movie ever, Love Actually?
Happy nerding. 🤓
Before you go ...
I very much enjoyed the conversations underneath my article on translation machines. It was great to see a lot of translators get involved as well. And thanks also to Else Willemsen, who came across a funny language error on her holiday in China.
(小 心 = xiǎo xīn = be careful; 落 水 = luò shuǐ = fall into the water).
Prefer to receive this newsletter in your inbox?
Follow my weekly newsletter to receive notes, thoughts, or questions on the topic of Numeracy and AI.