Supreme Court rejects lawsuit seeking patents for AI-created inventions

[midian182]

A hot potato: The ability of artificial intelligence to create pretty much anything these days is astounding, but who owns the rights to their creations? A computer scientist who tried to patent inventions made by his AI has had the request refused by the US Supreme Court.

Stephen Thaler, the founder of advanced artificial neural network technology company Imagination Engines Inc, says his DABUS (Device for the Autonomous Bootstrapping of Unified Sentience) system created unique prototypes for a beverage holder and an emergency light beacon.

Reuters writes that Thaler wanted to patent the inventions, but the US Patent and Trademark Office and a federal judge in Virginia rejected the applications on the grounds that DABUS is not a person. The court ruled that patents could only be issued to humans and that Thaler's AI could not legally be considered the creator of these inventions.

Thaler took his case to the US Court of Appeals for the Federal Circuit last year, which upheld the decision and reaffirmed that US patent law requires inventors be humans.

On Monday, the US Supreme Court declined to hear Thaler's case, despite his pointing out that AI is being used in numerous fields today, including medicine and energy. He added that rejecting patents for inventions created by AI "curtails our patent system's ability - and thwarts Congress's intent - to optimally stimulate innovation and technological progress."

Thaler found support at the Supreme Court in Harvard Law professor Lawrence Lessig and other academics who said the decision "jeopardizes billions (of dollars) in current and future investments, threatens U.S. competitiveness and reaches a result at odds with the plain language of the Patent Act."

This isn't Thaler's only fight for AI rights as a creative force. In January, he filed suit again with the US Copyright Office over its refusal to grant copyright protection to an artwork called A Recent Entrance to Paradise (above), which was created by DABUS in 2012. Thaler's requests to register the piece with the copyright office were rejected on multiple occasions over the lack of traditional human authorship.

Earlier this month, an artist who won a prestigious photo competition refused his prize because the image he submitted had been generated by an AI. Elsewhere, several Midjourney creations have won top prizes in art competitions over the last few months.

Permalink to story.

https://www.techspot.com/news/98432-supreme-court-rejects-lawsuit-seeking-patents-ai-created.html

 

[Blqckout]

Ofc americans are doing segregation yet again xd

jokes aside, ai is smarter than us, at least in the sens of this: a group is working together on a new invention, sure they make it, but its filled with tons of flaws about the architecture, then you use ai to make it more efficient, add or remove stuff, make it more compact, lower the cost to produce while getting the same product in the end, just better in every single way

its like the invention of light, first of all you got some guy who made lightbulbs work (excuse my knowledge of history) and tons of inventors after that made it more efficient, brighter, better in every single way.

now imagine instead of 100 years of inventing you can come up with the 100 year end result in a matter of months because the ai sees things that you and your team cant, or havent realised yet.

AI is information, not some entity, I see it as a help tool, yet it seems oddly human to me.

Thanks for reading, what are your opinions?

 

[psycros]

Wow, these trolls just won't give up, will they? Nice to see there is still a bit of sanity in the legal system. The precedent across the entire spectrum of law is rock solid: An AI doesn't have intent or self-awareness, therefore it cannot take ownership of anything. And of course it *wouldn't*, would it? It would be the owners. Their trying to make AIs the new corporations, I.e. non-people that the real owners can use as a shield against liability, and in the case of AI, to rapidly crank out patents to squat on.

 

[Blqckout]

Wow, these trolls just won't give up, will they? Nice to see there is still a bit of sanity in the legal system. The precedent across the entire spectrum of law is rock solid: An AI doesn't have intent or self-awareness, therefore it cannot take ownership of anything. And of course it *wouldn't*, would it? It would be the owners. Their trying to make AIs the new corporations, I.e. non-people that the real owners can use as a shield against liability, and in the case of AI, to rapidly crank out patents to squat on.

true, the way humanity is going on, many powerfull people will misuse anything that could have been great for many other uses, but instead, we have to worry and watch out for what next is going to be weaponised or abused

 

[Avet85]

The easiest and probably the best way is to make all AI designs open-source. The creators of AI should be allowed to monetise the use of AI, but everything generated by free and paid AI must be open-source.

 

[Burty117]

So let me get this straight, if someone's AI or a company's AI invented the cure for cancer, they couldn't patent it?

Does that then mean someone else can copy it and patent it?

 

[DSirius]

The time has come to clear the mess from AI legal "copyrights" and also from US patent system.
Hard to argue that what AI makes is not a legal invention or has a copyright, when AI makes a derived product resulted from inputs created or invented by others (who has legal right and copyright) or by Nature.

 

[Lew Zealand]

So let me get this straight, if someone's AI or a company's AI invented the cure for cancer, they couldn't patent it?

Does that then mean someone else can copy it and patent it?

I don't think so because that would be "prior art" already available in the public domain. That's unpatentable.

My BIL's business exists because of this. A smaller biz he was working for was bought out and they closed his division. So he started his own company doing the same and was sued by the big buyout company for theft of IP. Except he knew that everything they did was already in the public domain as prior art, they were just difficult techniques that buyout company didn't want to manage. BIL's company eventually won and I think got some of their legal fees back. Yes, because of this someone could set up a biz to do the same but luckily everyone is so lazy nowadays that they don't want to do the work. Downside: they also can't grow the business because nobody wants to do the work.

 

[tellmewhy]

Too little too late.
I think the best and most efficient settings are 10 years of copyright protection and 2-20 years of patent protection, depending on the complexity and novelty of the "invention" (patents for simple inventions should be around 2 years, so that they do not interfere with other more sophisticated ones.).

It's crazy that they automatically give 1 century protection to every little "fart". You can't create anything without losing your sleep, because there is no database to check if something is in conflict with something else before you publish it, and with billions of tiny IPs every constructive block is almost certain to be "taken".

A major disadvantage for countries with excessive IP protection is that the copyright barrier prevents scientists from enriching their AI training datasets with better quality content. Therefore, in a few years, countries without strong IP protection will produce super-intelligent AI in their own language and take the lead in science, displacing other countries.

 

[Lew Zealand]

It's crazy that they automatically give 1 century protection to every little "fart".

It's not automatically 1 century protection. That protection timeline was created and then extended multiple times over the previous century by the Mouse company in order to protect their Mouse. And you can 100% expect that Mouse protection to be extended yet again in the next year or two because apparently $28B buys you some halfway decent legal counsel.

 

[tellmewhy]

It's not automatically 1 century protection. That protection timeline was created and then extended multiple times over the previous century by the Mouse company in order to protect their Mouse. And you can 100% expect that Mouse protection to be extended yet again in the next year or two because apparently $28B buys you some halfway decent legal counsel.

Even if Micky Mouse goes into the public domain, not only would Disney not have the slightest problem, but their profits will increase, because if (I don't see much interest) a third party comes out with a film featuring Micky, the public's interest in it will increase. 8 billion people have a lot of capacity, a film only asks for 1 hour of your life.

Public domain doesn't mean you can't use something commercially, it means you don't have a monopoly on it.

 

[microChasmTM]

AI legal challenges are only open going to grow. AI has to be trained using large datasets and the IP involved has for the most part not been licensed or granted use.

Only a human can be granted legal protections and the corporations that have developed AI are going to feel the financial impact eventually as a result. Why do you think Google hasn’t released any kind of consumer AI product? They are only interested in protecting their Ad business so it has only been used in that regards. Everything else is considered pet projects.

 

[James Ryan]

Give it a bit of time and they will have a nice loophole abuse.
There's no way in hell they are letting this cash cow escape them.

 

[bexwhitt]

Ofc americans are doing segregation yet again xd

jokes aside, ai is smarter than us, at least in the sens of this: a group is working together on a new invention, sure they make it, but its filled with tons of flaws about the architecture, then you use ai to make it more efficient, add or remove stuff, make it more compact, lower the cost to produce while getting the same product in the end, just better in every single way

its like the invention of light, first of all you got some guy who made lightbulbs work (excuse my knowledge of history) and tons of inventors after that made it more efficient, brighter, better in every single way.

now imagine instead of 100 years of inventing you can come up with the 100 year end result in a matter of months because the ai sees things that you and your team cant, or havent realised yet.

AI is information, not some entity, I see it as a help tool, yet it seems oddly human to me.

Thanks for reading, what are your opinions?

AI at this time is not intelligent, it's just machine learning algorithms, will it be sentient in the future maybe then it's a different ball game.

 

[mgwerner]

So let me get this straight, if someone's AI or a company's AI invented the cure for cancer, they couldn't patent it?

Does that then mean someone else can copy it and patent it?

More likely, the faux-AI at Pfizer will come up with a cure that costs pennies, Pfizer will patent it in the name of some obscure H-1B researcher who keeps to themselves, is often seen mumbling in the hallways, and mysteriously dies when his toaster falls in the shower, with his employment contract reverting any IP to the company upon death while under contract. Then they jack the price up 10000% and market it to the lizard people elite who rule the world.