[Ethics] A heavy suspicion towards the Qwant search engine

hipeline

I was very supportive of the Qwant search engine. But a week ago, here's what the Ethics & Legal Affairs Officer [@QwantCom] Guillaume Champeau said on twitter and I don't really know what to think about this tweet that worries me considerably:

Against hate on the Internet, let us strengthen Justice by creating a "digital prosecutor's office" equipped with human and technological resources enabling judges to

to collect reports;

to obtain the identity as soon as possible;

to make an immediate appearance;

He later withdrew, stating this:

Tweet deleted because it is interpreted exactly the opposite of what I tried to say and Qwant's position which is obviously not to encourage censorship quite the contrary. I will explain myself more on my blog.

I'm circumspect.

LonM

@hipeline he says he'll explain more on his blog. It seems to me the best thing to do before casting judgement would be to find his blog post. I don't know where that is though.

hlehyaric

@LonM Here's the link to the blog post (in French):
https://www.champeau.info/blog/2019/02/21/defendre-la-justice-face-a-lincitation-a-la-censure-privee/

LonM

Looking at the blog post, the sense that I'm getting is that guillaume believes in:

To leave descisions on what constitutes infinging content to private sites rather than the government
Only involve the government / law in cases where obviously illegal content is being made
Private sites / giga-social-networks aren't doing enough at the moment to flag infringing content

This is what he wants / proposes:

Oversight from the justice system in cases where people are making illegal/hateful comments on social media
Involve humans in descisions as well as AI descisions on what content is "bad"
More transparency from big social media corporations as to how they go about detecting bad content

Having more of this, and involving humans in descision making, would be better as currently it is in the best interests of big social media platforms to just silently erase any content which might be illegal as deemed by an AI. You can see YouTube doing this a lot recently, using AI to make arbitrary desisions about videos. Guillaume argues that is bad censorship.

If I'm interpreting that correctly (which I might not be), that seems like a good descision to make. In fact, german regulators have already forced facebook to do exactly this: bring in human moderators to review content to find if it is illegal, as AI on its own isn't up to the task.

As for the original tweet of enabling judges and the law to act - from his blog post he says this is for in extreme cases "where there is no doubt of illegality".

hipeline

@LonM I don't want my post to become controversial.

Although he changed his profile just after, Guillaume Champeau originally spoke on behalf of his company.
Perhaps the deleterious political-media climate prevailing in France explains my hyper-sensitivity. I don't feel very comfortable being in the top 3 countries that get satisfaction from twitter for their request to remove tweets or twitter accounts (I would like to point out that I am not concerned): this indicator seems catastrophic for my country
The desire to abolish anonymity on social networks is the current trend in my country.
The ""Parquet numérique"" is the prosecutor, not the judge, the accuser in short.
Today the hate speech, tomorrow (in the upcoming elections) the fake news, and the day after tomorrow everything that will not be politically correct.

@hlehyaric Merci à vous pour le lien

LonM

@hipeline

He certainly shouldn't have been speaking on behalf of the company to begin with, if this is his personal opinion. That's not a good look.

As to the issue of how the French legal system handles online content, I must admit I'm not very knowledgeable about that.

I do think though that social media companies need to take more steps to moderate content. It might not have to be in relation to the laws of any particular country, but at things currently stand they seem to not really care about morals of any kind unless it looks like it will affect profits... At which point any semblance of freedom (of speech, of expression) goes out the window.

Having governments force companies to do more human moderation seems like a good thing, to me anyway.

hipeline

@LonM The impression of being caught in a pincer on one side by Internet giants who track all personal data and on the other side censorship which is more and more self-censorship, according to the panoptic architectural model.

Vivaldi is this essential link in the chain, to which a search engine that scrupulously respects privacy must be added. It seems particularly problematic to me that Qwant's Ethics & Legal Affairs Officer, in his private capacity, thinks exactly the opposite.

I knew that this "service" has received public support from French investment banker and current Macron administration, but I had not perceived the content of this junction.