Say No to AI Protect Act

Spoiler: One of the promises from the Artificial Intelligence Action Summit, the European Union is about to vote the AI Protect Act to face the AI crisis.

For some years, we’ve noticed that the AI sector has difficulties to find its business model. Despite massive investments, the companies lose money every years. For the human efficiency competes with AI models. Behind the buzz, some experts in the field admit, off the record, that the crisis is at a point where jobs are in troubles.

To avoid such catastrophic event, the AI Protect Act proposes a set of systemic measures and synergy moves by dealing with the heart of the problem: human beings looking at human creations.

Agency for Artificial Intelligence

The project was initially targeting a formal prohibition of content generated without artificial assistance (term replacing human in the law) and was condemned by the Pirate Party for this law attacks, as they say, the fondamental right of free speach.

Despite the support of the Greens at the parliament which oppose themselves against any attempt to artificial anything for the sake of biodiversity, it was the republican parties that bring a solution to the institutional dead lock by proposing a usage fee. A part of which will be invested in sustainable data centers.

The last project, which should be voted without any substantial changes (according to sources close to the case), propose the creation of a new European Agency for Artificial Intelligence. Being granted power of investigations and administrative justice. Before its creation, national agencies (ARCOM in France) will be in charge of collecting and distributing the fees.

The text also plan that the fee will be due and split between producers and consumers of those not-assisted content. The allocation will be fixed by order of the new Agency (and in the meantime, of the national agencies). And dissuasive fines are also planned for human beings pretending to be artificial beings.

the French Agency (Arcom) have already announced the creation of the label French AI Generated for content generating companies (such as ad agencies, publishing houses and the press). After an independent audit, study of the case by the agency, those companies will be allowed for simplified declarations and their presence in the official list of assisted information sources.

Social Networks

The big tech have claimed to want to respect the new rules and that they are currently working at new and innovative features for theirs users.

Linkedin said that the standard account humans will be able to only see (and be seen by) AI and robots. For thos feeling nostalgic, a new Premium authentic account will be available without those limitations.

After the fusion of xAI and X (ex-twitter), Elon Musk said that all content published by free users will be reformulated by its new AI so they are «free from the woke european imperialism». The premium users will be able to chose from a set of assisted reformulation before publication.

At the opposite direction, Mark Zuckerberg consider that AI is made to assist users in their daily lives by replacing the wall with friends’ content by a synthetic, AI assisted, review of those publications, following your preferences.

Off the record, an anonymous executive admitted: «there remain no human content for a long time».

Self hosting

All the indieweb sphere rebelled against this measure «took by brussels bureaucrats despite the advices of tech experts». We know the difficulties to determine the human ratio from their website statistics. But this time they are playing for the fee they must now pay.

Cloudflare and Google were excited to anouce the update of their captcha to meet the new context. It’ve been known for a long time that it’s more and more difficult to find a problem that only a human being can resolve (for the AI make such great progress to solve everything).

Both of them should shortly provide a new inverted captcha allowing any webmaster to identify robots by asking the solution of problem too hard for human beings (e.g. find where is Charly).

This race for weaponry between detectors and humans is promising for a hacker from Moldova gained access to those captcha and have written a Firefox module to detect and use an AI to solve them. The expert we’ve asked said that there is the same problem with age restriction to porn websites and the problem can not be solved at the hosting side.

Internet Providers

Complementary to collect the fees, the Internet Services Providers will have to take their part and have divergent point of view. Some spoke about the attack to the Net Neutrality (a convenient scapegoat), other see lots of innovative opportunities for their subscribers.

Following our sources, Bouygues and SFR should integrate some kind of content loging so they can emit precise receipt. Free is studying a all in one plan.

At its side, Orange, working with Arcom, will proposes an update of its hadopi labelled filtering tool, available since 2009, to integrate a white list of French AI Generated labelled website and thus propose a secure and free browsing for its subscribers.

By the side, the whole tech community is creating new ideas. We’ve been taught that Microsoft, working with the Ministry of cultural Affairs, to propose the same kind of functionnality in Windows 11 and Edge that will exonerate the citizen from any fee.

Online Petition

Even if the text should remain untouched until the vote at the parliament, the reactionnary force of the web still hope. The Quadrature du Net and Framasoft have launched an online petition for a revocation of the whole project.

Link to the petition