IBM invents printer that checks for copyrights IBM has filed with the US Patent and Trademark Office (USPTO) for a patent on a printer that checks for copyright violations. Big Blue's application describes a device that can check print jobs for text or images that have already been copyrighted, and then delete or block the infringing content from being printed. "The …
Quite so. I can't imagine why anyone anywhere would ever buy such a thing.
From the customer's point of view, if it actually works as intended it's no better than an ordinary printer; paper in, ink splatter, paper out just like normal. So, what'd be the point?
However, if (as seems highly likely) it's a screw up and deems your own picture of a sunset / nice view / picture of HM Queen Elisabeth to be oh-so-similar to this copyright picture over here in some other corner of the Internet when in fact it is one's own picture, then it's going to be a pain in the arse. Printing is annoying enough as it is without the goddamned printer taking it upon itself to not output parts of one's own page simply because some crummy program reckons you've half-inched it.
The only way something like this would be marketable if there was an overriding legal reason for companies to have them. Whilst printed image piracy is no doubt is going on, it's nothing like as bad a problem as online image piracy.
...so if this newfangled drm printer misses a few copyrights, does that mean the potential sueballs moves on to the provider of the equipment?
BTW: Is the printer manual copyrighted? Can I copy excerpts of the manual? Or will it be like a snake eating its own tail?
I have a Lexmark (i.e. ex-IBM) printer whose driver software one day wanted to install an update. Conveniently there was a button to click to learn what the update did. This revealed that its purpose was to "prevent the use of unauthorised ink cartridges". Surprisingly enough, I declined to install it.
-A.
"Still, it is not hard to see a practical application for the copyright-conscious printers in environments such as libraries or schools, where administrators would seek to limit the possibility of infringement from users wanting to print out partial or entire works."
Except that in UK schools, for example, there is a blanket licence to print limited extracts from copyrighted works, so the printer would need to know the length of the extract in relation to the length of the whole work.
That reminds me - about 30 years ago, our music teacher sent us to the library to copy a sheet of music. We had the phrase "One copy for the purposes of private study" drilled into us". I seem to remember it all went a bit Pete Tong when I was sent to get a copy for everyone in the class, thus triggering the guardian of the photocopier "It's one copy for private study, it's just there's a lot of us", followed not long after by a Music Teacher vs Librarian show down, which is a little bit uncomfortable, as they we're not very good at showing down. I put it down to lack of experience. I think technology would have made the librarian happier - "Computer says No" is a lot easier than "Thou shalt not pass". Music Teacher won IIRC
...in UK schools, for example, there is a blanket licence to print limited extracts from copyrighted works...
It is that way here as well. It's called "fair use." As this is a very common exception in academic settings, I cannot imagine there would be much uptake there, even if IBM gave these printers away.
Where I can see this sort of tech might be of some actual benefit is in corporate settings as part of a document management system. If you have restricted the ability to copy files to USB drives and blocked sending them out via email, it would be pretty lame for someone to be able to simply print them out on paper and walk out the door with them in that form.
If it is unviable then it is hard to see how the method could be "patent", in the original sense both of the word and the notion of patent protection. The fact that it is even possible to file this patent shows that the system is broken. As earlier comments have suggested, determining whether you have the right to print a document is clearly so far beyond the capabilities of even humans, let alone artificial intelligence, that this is the moral equivalent of patenting a perpetual motion device.
This is why if you're not making anything and profiting from a patent you hold, is should be torn up and thrown in the bin. Preventing someone else from doing something you don't want to / can't be arsed to is monumentally pointless.
Patent applications should be accompanied by a working prototype. They should then be reviewed periodically to asses their ongoing utility.
In the future it'll be illegal to own a printer that doesn't come with a Copyright Infringement Prevention device. Besides it sounds like great way of spying on the population, when everything you type is sent back to Minitrue head office. The Staatssicherheit could have only dreamed of such technology.
Don't be so sure.
There is prior art.
Try printing dollar or Euro bills on a high class color laser printer (something that can print a realistic enough replica) or try to copy them on a color copier. They probably no longer do that on the low end SOHO kit, but the high end proper office kit still has it.
Puh-leeze, Mr Hand. That's *one* document, precisely specified and stored locally on the device. If IBM are contemplating building a printer that already contains all possible copyrighted document images then I will eat my words, my hat, and probably all my toner cartidges just for the LOLs.
It isn't "one document". First, it's more than a single currency, on our Ricoh, it does it with several, like US dollars, and others.
Then, the Euro bills were recently updated, their picture and color being substantially modified, and yet, it still recognizes them, without an update.
It isn't "one document". First, it's more than a single currency, on our Ricoh, it does it with several, like US dollars, and others.
Then, the Euro bills were recently updated, their picture and color being substantially modified, and yet, it still recognizes them, without an update.
That's why I use proper intaglio printing for all my counterfeit needs.
"Then, the Euro bills were recently updated, their picture and color being substantially modified, and yet, it still recognizes them, without an update."
That's because the almost invisible yellow dot pattern the scanner/printer is looking for has not changed. Google it and be amazed.
Still, it is not hard to see a practical application for the copyright-conscious printers in environments such as libraries or schools...
It's also not hard to see the practical application of vans with air-tight compartments at the back into which the exhaust pipe, rather incongruously, opens.
That doesn't mean that any of these are particularly good ideas.
(As an aside, said vans were apparently a Soviet invention, though history does not record whether they were patented, although it DOES record that the inventor, one Isay Berg, got purged by the purge he himself so enthusiastically supported)
I know this happens all the time, but allowing a patent on a vague idea, rather than an actual working device seems to be absolute lunacy. - even with pretty pictures this is no more than a pipe dream.
"Hey guys - wouldn't it be good if we had a machine that....." *
At best this sort of patent is bonkers.
At worst it allows someone can't or won't to block someone who can design a device.
Imagine telling the first 'plane designers that they can't do that because "we already thought it would be a good idea and got the patent. No, we don't know how, but that's beside the point."
*Irrespective of whether this copier is a good or bad idea.
The moment lawyers became involved in the patent process it was the beginning of the end of real patents where you had to have a working model to support your claim.
The patenting of ideas is a travesty being lead by the US patent office backed by lawyers and marketing droids and because of this it should be disbanded or, at least, returned to the issuing of real patents backed by actual working objects.
Clicks icon which converts text to image, then prints it.
The USB cable grumbles about having to carry extra bits.
"Oh darn, this image won't print."
Inverts colours on image. Inverts colours in print driver. Then prints it.
This would be as effective as disabling the right click Copy option. Which is to say, not at all.
Remember that most scanners/copiers sold in the last few years already have similar (if much more constrained) technology in place for identification of banknotes (the so called EURion constellation) - and they managed to get this rolled out behind everyone's back without even acknowledging its existence.
Fair point on possible secrecy, but regarding feasibility EURion is the business of recognising one (or few) known images that are designed for easy recognition, versus trying to recognise everything under every circumstance. Of course if an unstable best effort is good enough than here's "return (bool) (rand() & 1);" and we'll go to the pub already...
Here's the real meat of the copyright verification, from Fig 2 where the "copyright infringement prevention program" is laid bare (in a goddamn flowchart):
#216 - Identify potential copyrighted materials from file
#222 - Search online resources for identified potential copyrighted materials
And that's it! All the complexity of doing this (imagine even describing to a smart human how to do this task, let alone making a useful algorithm for it) blithely ignored. In fact there are about six other paragraphs revolving around the "revolutionary" idea of looking for a match in a cache of previous evaluations, i.e. "prior art since forever", i.e. "the only bit of this idea we understand well enough to have a hope of actually doing"
It's truly "underpants gnomes" stuff. The best explanation I have is that IBM prides itself on the patent output and provides big incentives to its staff ("have some cash" or maybe just "keep your job") , so the patent mill is cranking and gold or fools gold, in it all goes...
Mongo "All the complexity of doing this (imagine even describing to a smart human how to do this task, let alone making a useful algorithm for it) blithely ignored. "
At a place I used to work at, we called this "design by waving your arms around". As in sitting in the design sessions and the person up the front with the whiteboard marker would just make a sort of whirly-hand gesture and draw a swirly box.
Icon cos that's what this technique normally did to delivery schedule.
Set the paper size to "US Letter" (in Europe) or "Japan M4" (in USA), then either (a) simple user gives up in frustration or (b) sophisticated user is beaten to death by horde of simple users when they go to the halted printer to clear the error
That doesn't work too well if using a Postscript printer set to automatically resize.
It also doesn't work with some PCL printers that also support page resizing.
There is also differences to be considered between Page printing and Document printing.
Inkjets are not used in most offices nowadays but when they are they are usually for more specialist work. Inkjets have a far wider colour space than laser printers.
But inkjets on the whole dont print as page or document printers but as line/print buffers.
As regards copying bank notes it is one of the reasons that early colour laser copier machines did not support double sided printing. They also had the ability to remember in service codes when and what was scanned in and the engineers working on them were duty bound to notify the authorities if they found one being misused.
That was in the days when the machines were extremely expensive. Nowadays they are almost 2 a penny in comparison and the technology in them has progressed considerably.
At one point of time EFI (Electronics For Imaging?) earned more money per employee than any other listed company. So they made their money from the systems even if the user couldn't!
"They also had the ability to remember in service codes when and what was scanned in and the engineers working on them were duty bound to notify the authorities if they found one being misused."
Yer what? Back when I "did printers", I was never informed of that or asked to check "the logs" or peep through the HDD contents to see what our customers were up to. In fact, doing that would be strictly verboten unless there was a very specific reason to examine the HDD contents. If it was an MoD or other government site, the HDD would normally be removed and retained if the printer was to be taken off site and a full factory reset performed. In some MoD sites all the PCBs had to be removed and destroyed on site if it had been used in any restricted areas in case any "secret" data had been hidden away in NVR or EEPROMS.
As seen in a previous On Call article, service engineers should NEVER look through a users data unless specifically instructed and authorised to.
"Many printers won't print currency."
You know that, and I knew that.
Till a few minutes ago I didn't know how it was done, and frankly couldn't see how it could be done via the usual image recognition methods (which are presumably part of this IBM patent, though it doesn't actually say so).
Now, courtesy of Dave Harvey's post a few hours ago, I know that currency recognition doesn't use image recognition as we know it. It uses a not very well known agreement between currency designers and high end printer companies, to embed specific recognisable patterns in non-copyable currency designs and then copiers refuse to copy anything containing those specific patterns (oversimplified).
See "EUrion constellation".
Thank you Dave. Have a pint.
So the printer will send enough of every document you print, probably unencrypted, to IBM (or a 'carefully vetted partner') so it can be checked for copyright. Self contained DRM won't ever work because I could print a text file, or a bitmap, or a PDF already in circulation, so un-watermarked, but still copyrighted.
I imagine the printer wouldn't work, or would fall back to very low quality slow output, if your internet connection was flakey, or IBM's servers were flakey or anything really. This is beyond idiotic, right up there with Windows 10's telemetry uploads or it deleting applications Microsoft doesn't want you to have. I await version 2.0 that will inject adverts into your documents, and bombard you with spam.
I don't get what's collectively wrong with the people in these large corporations. Presumably sociopathy is contagious, starting at the top and spreading downwards.
Who is going to manage the database?
Also how does it manage:
Geo variations
Licences
Fair use
copyright holders
This is unworkable.
No printing allowed if internet connection goes down?
Privacy: Snooping on what people are printing. Iran, NK, Saudi Arabia etc will love it.
This post has been deleted by its author
This patent might describe a so-called method (and not simply an idea, which is not patentable) but I'm still amazed at how people come-up with these "methods" that someone from middle school could have thought of for a pet project and yet have no working prototype to at least validate the basic claims.
Heck, I could "invent" a method for hitting the nails with the side of the hammer because that way there is more surface area and you are less likely to miss... is that patentable??
And in addition to that, why are corporations always trying to "avoid" courts by implementing their own legal systems (see Youtube policies or other DMCA examples defaulting to "guilty until proven innocent" instead of getting a fair trial).
What's more interesting in either case is that lawyers are stretching the semantics of the laws and judges/jury find it more difficult to apply common sense and instead have to fight against many technicalities, not to mention disproportionate examples like mall cops shooting someone in the leg for not paying a $1 energy bar.
@"Gee, wouldn't it be amazing if a company could put technology in place to protect musicians from having their work stolen"
You do know Google's 'Content ID' don't you? Googles automated take down system that does exactly what you're asking for.
https://support.google.com/youtube/answer/2797370?hl=en
So if this is a networked printer, will it squeal to the copyright claimant about the attempt to copy their work?
And will it, in the process of examining the copyright, hand company data across the network to outside?
If IBM doesn't trust its customers not to violate copyright, why would their customers trust IBMs kit to act in their best interests?
> All i do is put © Roger Kynaston 2016 at the bottom
I think that's not quite right. "© Roger Kynaston" is a notification to readers of your claim, not evidence that you hold copyright. If you create something (poem, picture, program, etc.), copyright is implicit; that is, you hold the copyright without saying anything to anyone.
Proving it is a whole different matter, of course.
Once you've got this, it should only be a little more work to prevent people printing libellous statements. There are quite a few countries where criticism of the monarch / state / religion is illegal, it could check for that too. We've got anti-bullying and anti extremism laws now so it could check for people printing mean things or leaflets for banned animal rights groups. Many of those examples actually have longer jail sentences than copyright infringement so you could argue that preventing them is a far more worthwhile activity than checking copyright.
Or, you know, we could not do that because it's a horrifying idea.
Doesn't this need a local copy of everything in the world that has ever been copyrighted, in order to be able to do the comparison to check?
Assuming that it's possible to get copyright permission to keep electronic copies in the printer, what bandwidth would be required to keep it up to date?
Considering that much of the info I look for is behind paywalls with ludicrous per-page access charges, how does the maintainer of the database afford the copy needed to check against? I do understand that companies in the business of collecting rents on our every waking thought often have "bulk discounts", but you would have to have something like a 99.999% discount to be able to effectively block the fair-use of snippets from just one Scientific Journal.
If the database is off-site, and ignoring the "how do I print if the internet is down?", what does a lawyer or doctor do to maintain legally mandated confidentiality (also a problem for Win10 users, but another day...) Obviously not a problem as long as use of such printers is optional, but we know how long that's going to last.
Just as the DMCA was the death of due process, allowing private parties like Google to act as judge, jury and executioner when it comes to removing the fuits of free expression like a 14 year-old's video game review, thus sweeping aside almost a thousand years of Western jurisprudence, this invention will mean the end of fair use -- and so the end of free speech and democracy. But then IBM isn't new to the business of enabling dictatorship and destroying human rights, are they? Their tech was so helpful, after all, in the effort to ennumerate a particular ethnic group in Nazi Germany, just before their customers embarked on the second phase of their project: the mass murder of those whose details were so carefully recorded on IBM's cardstock. This is what unaccountability leads to.
When if /I/ own the copyright? I try to print some of my work, my printer refuses because it's a copyright work (as almost everything created in the last 75 years will be - copyright on written work lasts from death of creator plus 75 years). How do I tell the printer to do as its damn well told and just f****g print the stuff I want it to print?
And what about the works that I hold the copyright on, and my license for that is "use it as you like without usurping my copyright"? Remember, ****EVERYTHING**** is copyright, just by virtue of it being created.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2024
