Fear the Future: Art, Tech + the Magnus Scandal

If I've learned nothing else of real-world value from horror movies, I've learned that what seems too good to be true often is. Your careerist spouse suddenly puts aside their burning professional ambitions one evening and decides to start a family with you? Well, it might be a one-night only gear-shift in a black-magic pact that leaves you rearing Lucifer's child. You just built the perfect new home in a gorgeous allotment thanks to your plush job with the real-estate developer? There's a chance its foundation rests on a Native American burial ground, and a ghost-infested TV will drag your daughter into the netherworld. You're able to download an app that claims to "democratize" the art market by instantly supplying prices for any work whose photo you upload from a gallery, museum, or art fair? Chances are it'll be out of circulation in a few months, after a slew of other businesses accuse its maker of stealing their copyright-protected data to generate those magical results.

That's what what happened over the weekend, when Apple yanked would-be serial disrupter Magnus Resch's namesake software, Magnus, from the app store, after a trio of German galleries joined the growing ranks alleging that Magnus's database appropriated some of their images and pricing without permission. Art-market analytics firm Artfacts filed a complaint to Apple in April reportedly showing multiple cases where Resch's app recycled its paywall-protected data word for word, digit for digit––including errors. And around the same time, Artsy also claimed Magnus infringed on selected photos provided by its gallery and dealer clients, bringing the app's alleged victim count to five. 

As of my writing, Resch himself remains defiant about the data-theft charges, in both actions and words. While new users can no longer download the app, it still seems to be operational for anyone who adopted it before this weekend (at least, based on my personal testing). Resch has also countered to the press that any data-driven entity accusing him of wrongdoing simply doesn't like being beaten at its own game. To emphasize the point, the Magnus website now opens with this pop-up:  

In other words, you can't make an omelette without breaking a few eggs, right? 

The issue, though, is that you most definitely CAN make an omelette without raiding your neighbors' henhouses. It's just a lot harder to do.

That said, it's not entirely clear to me at this point whether or not Resch has actually been hacking through chickenwire––and, even if so, whether he's been doing it on a grand enough scale to undermine his entire venture.  

First, Artsy has already confirmed that Resch complied with its takedown request months ago by removing the offending images from his app's database. That reduces the number of still-aggrieved parties to four. And while I'm not an intellectual property attorney, it seems to me that, if its claims are accurate, Artfacts makes far and away the strongest case that Magnus incorporates any actual and verifiable improprieties. Data is Artfacts' entire business. That's why only paid subscribers have access to its market info, which in turn implies to me that distributing any of said info to the non-paying general public represents legitimate infringement.

The three German gallerists look like an entirely different matter to me, at least based on the details I've been able to find. One of the three, Cai Wagner of Wagner + Partner, claimed on record to artnet News that someone from Team Magnus acquired the supposedly sensitive info in question by lying about his identity and intentions in an email inquiry to which the gallery voluntarily responded with images and pricing. This is not exactly a novel way to ferret out potentially useful data in the industry. Gallerists, dealers, and advisors do it to each other all the time. The only difference is that Resch (allegedly) went public with the goods rather than keeping them for his own private professional use.

Did Wagner's brothers-in-arms fall for the same ploy? And if so, does using that ploy qualify as a punishable offense? I'm not familiar enough with Germany's business standards or the app store's rules and regs to say. If it happened here in the States, though, I would argue that Resch violated the informal gentlemen's code that governs much of the art industry, but no more than that. And I'm not convinced an etiquette breach justifies keeping the app off the market, especially if the breach occurred because the gallerists handed out what they saw as sensitive information to someone they didn't bother vetting first.  

Still, let's suppose that all of the outstanding data-theft claims against Magnus are valid. That leads to the second question, which is much more germane to the app's future: How much stolen data are we actually talking about?

Only the people directly involved in these disputes know for certain. But if three of the offended parties are in fact single-location German galleries who coughed up data based on manual email inquiries, then Wagner and his cohorts can't be waging war with Resch over more than a handful of pieces in Magnus's database. No reasonable gallerist would ever offer more than that to an unknown "curator" soliciting info remotely––and, in the current landscape, any gallerist who would deserves to be exploited by competitors. (To quote NBA analyst Jeff Van Gundy, dumb gets you beat.)

That leaves Artfacts as the great unknown. But here too we can at least estimate the alleged breach's rough order of magnitude. Magnus advertises about eight-million auction and private-sales results. Artfacts advertises about 30,000. Even if Magnus committed grand theft data by making off with Artfacts' entire library of transaction records––which I've seen no indication of, to be clear––it would still represent only a small fraction of the app's overall content. Magnus would be perfectly capable of continuing on as a valuable resource without them, assuming its other ~7.97 million price points are legitimate, 

Yet that assumption is the key to this entire saga. Artfacts CEO Marek Claasen argues that, "There was no way [Magnus] could have amassed that much information so quickly. The kind of data the app offered has to be collected and organized by hand." I had a similar reaction when Resch unveiled the app five months ago, with no advance warning. But Melanie Gerlis's piece for The Art Newspaper (linked earlier) reports that Resch has kept a team of roughly 80 researchers on the project since 2013, while the app has also been sourcing additional data from cooperative industry figures along the way. Is that a big enough team to compile eight million trustworthy results in less than three years, without resorting to grand IP larceny? I honestly don't know. (Though, as usual, I'm skeptical of everything until I see hard proof to the contrary.)

In the end, it may not matter. The lingering presumption of foul play strikes me as the biggest problem with the Magnus fiasco. In general, art-industry insiders already view technology and transparency the way a father views the frat guys checking out his freshman daughter on move-in day. And why wouldn't they, given that every private dealer's business relies on weaponizing information asymmetry, while art valuation itself is essentially an inside job? So the current accusations against Resch only reinforce the sector's existing prejudices against innovation as a whole.

Resch himself only worsens the problem. Remember, he's already alienated many in the industry (myself included) with some of his previous efforts, particularly his frequently savaged book Management of Art Galleries. The man can try to portray himself as the people's champion if he wants, but that's a dubious costume to wear in a market where the very wealthy effectively qualify as the only paying customers. If Resch becomes the figure most associated with an industry innovation effort, that's a terrible outcome for technology in the arts. You don't want your mascot to be Robin Hood when the only people living in the city are aristocrats. 

Unless Resch can prove without a shadow of a doubt that Magnus exclusively runs on cleanly acquired data, then, we risk entering a future where everybody loses. Established art insiders only strengthen their already-historic resistance to innovation, which continues to stunt the industry's overall development. The less privileged members of the arts community lose a promising inroad toward transparency, which further discourages their participation and, as I've argued before, actually hurts business in the long run. And the art industry, like both science and my favorite horror franchises, only gets to advance one funeral at a time. Personally, I can't imagine an ending more disturbing than that. 

*UPDATE: For those interested in an informed American legal perspective regarding Magnus's potential violations (or lack thereof), longtime reader Brian L. Frye, an associate professor specializing in copyright and intellectual property at University of Kentucky Law, had the following to say: 

"For better or worse, the copyright issue is probably a red herring, at least under US law. Facts are not & cannot be protected by copyright, so Resch didn’t infringe anything by copying the data. Using images is a closer question, although the Supreme Court has held that using thumbnail images to identify search subjects is a protected fair use. Resch could be in trouble for violating terms of service, a la Aaron Swartz, but that’s not clear from any of the articles you posted.

"The canonical case on the subject is Feist v. Rural, in which the Supreme Court held that copyright does not & cannot protect “facts,” essentially for somewhat inchoate constitutional reasons. As a consequence, copyright can protect a particularized expressions of factual information, but not the information itself, and if the information can only be expressed in one or a limited number of ways, it can’t even protect the expression. Feist v. Rural said that copyright couldn’t protect the white pages listings in a phone book for that reason. The one possibility would be if the Artfacts price data involved interpretation, as courts have held that the information in auto “bluebooks” could be protected, as the price estimates involved judgment. But that would still only prohibit wholesale copying & pasting & would only apply to the price estimates, not objectively factual information about the works. Which honestly, price estimates arguably are. Or at least, what if the copier considers each estimate & concludes it is correct?

"The Terms of Service issue is not about copyright at all, but a contested internet law issue about whether internet service provide can & should be able to control how you use information you obtain from them by requiring you to agree to terms of service before you use their website. In other words, Artfacts could say, by buying a subscription, you agree not to share the info we provide in certain ways. Doing so could be a contract violation & potentially even criminal, under the Computer Fraud and Abuse Act. That is really not my area & my knowledge is limited, but the statute is written REALLY broadly, so who knows what they could do with it."