European Parliament Prepares Tough Measures Over Use of AI

The European parliament is preparing tough new measures over the use of artificial intelligence, including forcing chatbot makers to reveal if they use copyrighted material, as the EU edges towards enacting the world's most restrictive regime on the development of AI. Financial Times: MEPs in Brussels are close to agreeing a set of proposals to form part of Europe's Artificial Intelligence Act, a sweeping set of regulations on the use of AI, according to people familiar with the process. Among the measures likely to be proposed by parliamentarians is for developers of products such as OpenAI's ChatGPT to declare if copyrighted material is being used to train their AI models, a measure designed to allow content creators to demand payment. MEPs also want responsibility for misuse of AI programmes to lie with developers such as OpenAI, rather than smaller businesses using it.

One contentious proposal from MEPs is a ban on the use of facial recognition in public spaces under any circumstances. EU member states, under pressure from their local police forces, are expected to push back against a total ban on biometrics, said people with direct knowledge of the negotiations. Agreement between MEPs, who have been fighting over measures to police artificial intelligence for close to two years, is critical to kick-starting broader negotiations over the AI Act. The proposed law would represent some of the toughest rules on the development of AI and comes in the wake of rising concerns about potential abuses of the technology.

Sigh time to go back to a dumb phone

[thegarbz]

I won't be able to unlock my phone anymore in a public space.

Re: Sigh time to go back to a dumb phone

[FlowVoid]

Don't worry, you can still unlock your phone. The regulation applies to "remote" recognition systems: "notion of remote biometric identification system as used in this Regulation should be defined functionally, as an AI system intended for the identification of natural persons at a distance through the comparison of a personâ(TM)s biometric data with the biometric data contained in a reference database, and without prior knowledge whether the targeted person will be present and can be identified"

MAD 2.0

[Monolyth]

Moratoriums and laws limiting machine learning and what we are improperly naming "artificial intelligence" are a joke. Not everyone can make an "a bomb" because it requires materials that can be regulated worldwide. A smart enough individual has all l they would ever need in the Internet to make an "AI bomb".

"reveal if they use copyrighted material" - YES

[itsme1234]

Broadly speaking mostly everything is copyrighted, including but not limited to Wikipedia and Linux and the page you're reading now. Sure, there are exceptions like works that went to the public domain and of federal government and the digits of Pi and and and but you probably won't be making much of a useful training out of those.

Re:

[airport76]

As usual, the quality of the reporting is more than questionable. The word "copyright" only appears once, in the introduction, and does not appear a single time in the proposal, as you can verify here [artificial...enceact.eu].

More information: https://artificialintelligenceact.eu/ [artificial...enceact.eu].

Re:

[Zak3056]

Sure, there are exceptions like works that went to the public domain and of federal government and the digits of Pi and and and but you probably won't be making much of a useful training out of those.

That's an interesting view to hold, given that a traditional liberal arts "classical education" needs no materials that have not been in the public domain for decades, or even centuries. Also, in the EU, the works of anyone who died in the early 1950s are in the public domain, and, of course, that line moves forward yearly.

I think you are severely underestimating how much information we are talking about, here, or the utility of that information if you're interested in things other than making memes, or ha

Nothing to do with content creators

[VeryFluffyBunny]

Re: "a measure designed to allow content creators to demand payment." - Typically, it's not the content creators who demand payment, but rather the litigious, hyper-exploitative, unscrupulous media corporations that claim copyright ownership over anything they can get a away with. The system is already stacked in their favour and they want to skew it even further. WIPO needs to be reigned in, made transparent, and opened up to democratic oversight instead of being co-opted by corporations.

However, that s

Hm, is that the same EU that's always talking abou

[unami]

needing some European Google or Facebook? Not going to happen if they're hindering developement because they are panicked because of a glorified autocomplete. Also, do I as an artist get to pay royalties to the material I trained with? As an EU citizen, I'll probably use a VPN to access AIs then.

crooked road

[Micah NC]

If an art student studies commercial, proprietary magazine covers to learn how to draw models, that's fine.

When an AI does that, oh, we have to bring technical innovation to its knees.

EU citizens will suffer for this.

Re:

[nomadic]

For a very unorthodox definition of "suffer."

EU doesn't understand tech.

[bool2]

----BEGIN RANT----

The same EU that gave us those pointless and distracting cookie popups that appear on nearly every web page. The ones that place a huge, continuous, collective burden on the consumption of information. The ones that break screen readers and translations - all in a half arsed attempt to protect "privacy."

We value privacy. But we need our government organisations to come up with even half-way competent solutions instead of playing senseless power games with big-business.

This premature contro

Ideology

[Local ID10T]

This is an example of why the EU does not lead in technology. It is not a lack of talent/ideas. It is the restrictive legal framework.

If you limit the development of technologies, they will be developed elsewhere by people who do not share your ideals/morals.

Re:

[account_deleted]

Comment removed based on user account deletion

So, maybe Brexit wasn't such a bad idea

[mkwan]

I still think Brexit was a stupid decision, but this is the kind of thing the pro-Brexit camp was talking about.

Britain might now see a wave of AI startup refugees fleeing the EU. DeepMind is already based there, so they're got the basis of an ecosystem.

And that's good...?

[HannaJord]

It is important to note that such regulations aim to address concerns around privacy, transparency, accountability, and ethical use of AI, while also promoting innovation and competitiveness in the field. Nonetheless, there may be various perspectives and trade-offs to consider, and the final outcome of the proposed regulations will depend on the negotiations and feedback from various stakeholders.