I think it's a completely reasonable position that companies making self-driving cars and question/answer systems are legally liable for any errors.

But if you hold that position, you also have to be fine with companies not offering products and services in your country. AI systems will eventually be good enough (in 10-20 years) for companies to be able to deploy such systems with sufficient accuracy to afford the lawsuits. Until that time, such countries would just not have access to systems before they were bulletproof.

at the scale that google is at, it doesn't make an ounce of sense to hold a company liable for a potential mistruth. what if there's a 1/1000 chance of some error, then the company could be sued millions of times per day.

down vote all you want, but I firmly believe this is an example where the user needs to use some judgement on the information they receive and have some critical thinking skills. google would be right to remove all AI results from germany.

It doesn't "bury [it] deep in their TOS", it says right under the box:

> AI can make mistakes, so double-check responses

> Same for self-driving. Your car is not self-driving until it accepts liability and you count as just a passenger

Mercedes-Benz does this in limited cases. Waymo does it generally. (In China, Level 4 and 5 transfers risk to the manufacturer. This is the correct way to do it.)

Banning all technology because someone might misuse it is an illogical extreme.

As far as I can tell the ruling is more nuanced. If AI is defaming you, there needs to be a way to correct the record.

A company being open to liability does not mean it is always liable, just that it can be if it really messes up (especially if there are aggravating circumstances, e.g. you need to drag them to court to issue a correction).

Why would that mean AGI? You can get into liability-accepting territory by restricting scope, a lot easier than by making your AI smarter.

Self-driving cars don't need to be particularly good for companies to make models where they accept liability in some circumstances, and the cars refuse to drive in other circumstances.

Ai results that nobody wanted in the first place?

> The true mark of AGI

Can we just trash this as a marketing term? If/when AGI arrives there will be no point quibbling over competency. What we are looking at is just bad search results

Why is it good? Everyone with common sense knows AI can be wrong. And it’s not buried in their TOS. It’s in the chat box. But even if it wasn’t, it’s ridiculous to create liability for AI chatbots.

Nearly the entire American tech industry has been super heavily selected for people who undervalue the legal language with crazy implications buried everywhere.

Otherwise most of it would not even exist.

Everyone would have continued paying out the nose to the IBM’s of the world year after year (who had unusual willingness to sign short ambiguously worded custom contracts to their own disadvantage, if paid vast amounts of money).

And be on mainframes to this very day… maybe Y combinator and HN wouldnt even exist in that world.

> But watch as Germany soon loses AI Google results.

Oh no! Anyway ...

Of course, they know this. The entire point is be able to rewrite people's awareness.

Should we hold scientists and journalists liable if they say false things or misrepresent things?

You'd think so, along with other countries that have defamation laws. But there's no indication of any penalty, and Google wasn't even made to pay all the legal fees. Perhaps their business model (if there is any) can cope.

that would be hela curryketchup nice

It depends if Google feels the profit is worth the risks.

What profit? I don't know either but they enabled this for a reason right?

Due to costs of running frontier models on every search request, Google simply does not: the failure rate is so high when you are just expecting an objective output.

Imagine a search for your name resulted in an AI summary saying you are involved with child-trafficking because low capability model linked your first name and perhaps a couple articles on supporting children non-profits to it — and then offering that in a convincing sounding summary right at the top!

In normal flow yes, but likely not if you intentionally entrap it to say something wrong.

A disclaimer and couched language will probably fly through. And it's going to matter what expectations an user could reasonably have, too.

That’s not very typical in software. Especially software you don’t pay for (with money)

That is exactly the point of the ruling, though... they are saying that AI summaries are NOT the same as search. If Google was just returning search results, and then users clicked on a website and read the content there, Google is not responsible for the content.

If instead Google gives you an answer right there on google.com, without going to another site, they ARE responsible for it.

That makes sense to me?

That's the difference between returning search results and interpreting the information and summarising them. If a newspaper says 'so-and-so has been arrested for theft' it's not the same as them summarising to 'so-and-so is a thief', the second is potentially libel. Why should Google be held to a different standard?

The title is misleading IMO. It should say "German ruling declares Google liable for libel in AI Overviews"

I was prepared to say the same thing as you but after reading it seems totally fair.

The key difference is that this would be illegal if a human wrote it too.

Google itself is more trustworthy from a normal person perspective as they use it a lot.

None of "AI" companies call their apps "Entertainment fun text generator". They are call them serious names, use words like "intrllegence" and "thinking".

So yeah I'd think if any of "AIs" start to recommend to drink some bleach or take a flight from a 10th floor window these companies should be liable.

I think it's very clear that Google's AI overviews go far beyond just searching Google because they often incorrectly compile sources to come up with an incorrect answer. For example of this look at the comment I made in this thread

I go to Google to search, but get spammed as if I wanted to talk to a chatbot (and a very poor quality chatbot at that).

This is a gigantic own goal for Google. The average person’s impression is that Google AI is much worse than ChatGPT, even though that’s not actually the case. Google is shoving a terrible model in everyone’s faces.

Playing the perception game wins you the perspective price.

Nothing is free. Google benefits off you when they show you search page. Either today (ads) or later

> (in spite of very visible disclaimers, ToS, and usage policies)

If you sell food, in a food stall, labeled as food and you add a disclaimer that it is toxic and will make you sick. You are still selling toxic food and you are liable for it.

Google is pretending to give answers to your questions. They offer you a service about answering questions. And then they add a disclaimer "we do not answer questions just write bullshit". That is still fraud and Google should be liable for it.

> isn't this just a soft ban on the deployment of non-deterministic software?

Tetris is non-deterministic and it is not banned like millions of other programs. I do not follow you.

LLMs are deterministic, they are only non-deterministic when you add a temperature.

Where did it say the liability only applied to non-deterministic software?

no, just a ban on using non-deterministic software for situations where deterministic responses are expected.

What to do if the software automatically and wrongly libels you on a public search engine?

Honestly I can understand the ruling, but the side effects might be severe.

If people do say that, it's a true statement and thus fine. You are allowed to report that regardless of the truth of x/y/z/a/b/c

The instance of this ruling people apparently did not actually say any of the offending claims. 'The court also found that the AI overview made claims "that are not even made in the search results." None of the linked sources drew any connection between the plaintiffs and the shady companies the AI mentioned. The court called these "the defendant's own statements."'

Certainly, if this is pointing to the actual pages where the actual people express these things. Otherwise that's equally unfalsifiable claims, could be completely made up or actual truth.

One way to formulate things that would be less would be "once support a time, in some fabulated world, it's not impossible that some imaginary character would say something following some reason." But then, of course this is not aligning the the deception scheme pushed by companies putting in their interface that the "machine is thinking hard for you".

One that doesn't maim/injure/kill you is a pretty good standard. And before you call bs, look at all the foraging and chemistry books that are for sale on Amazon that are AI.

> Pretty much the exact thing that Google AI Overview does.

No, the article implies the court’s logic is that the AI search results are presented as search results and that’s a big part of why they are liable. It seems like the court (again, according to the article) does not find the disclaimers that Google has slapped on the AI results compelling because again, it chose to represent these as a summary of search results and it is aware of the failure rate.

> The court also found that the AI overview made claims "that are not even made in the search results." None of the linked sources drew any connection between the plaintiffs and the shady companies the AI mentioned. The court called these "the defendant's own statements."

> Google built the AI, Google offered it to users, so Google owns what it produces, "because it alone has influence over the AI's offering and the algorithms with which the AI operates."

Google does not, as a general rule, control the actual content of search results, but usually there’s a distinction between the ranking and presentation of the results vs. the actual content. In this case, the court is basically saying, “You sold this to people as a search summary, you know it might be full of crap and you chose to do it anyway. No, you don’t get to claim the equivalent of a US safe harbor defense.”

I don't see any reason why it wouldn't.

It requires the claim to be made with "willful disregard for the truth". Notifying someone, especially with a cease-and-desist on fancy letterhead, makes it legally clear that they know better, and thereafter would be definitely libelling you (assuming the claims are in fact untrue and harmful). But you can still sue them for the claim prior to the notice, you just have to prove they should have known better prior to making the claim.

In this case it looks like they were notified and didn't do anything.

That's just America.

Who will adjudicate private disputes between citizens if not the state?

> Does this extend to ads displayed in search results?

Probably not, for the same reason search results aren't an issue.

mr prime minister, we cannot allow an "automatic libel machine" gap!