On June 17, 2004, Cory Doctorow walked into Microsoft’s Redmond offices to argue that one of the technology industry’s major investments was fundamentally misguided. Microsoft and other companies were spending heavily on digital rights management, or DRM: systems designed to control how digital music, films, ebooks and other copyrighted works could be accessed, copied and played. Doctorow’s message was not that Microsoft needed better DRM. It was that the entire project rested on the wrong technical, legal and economic assumptions.
Doctorow approached the subject from an unusual position. He worked on copyright issues for the Electronic Frontier Foundation, but he was also a professional science-fiction writer whose livelihood depended on copyright. That allowed him to reject the easy caricature that opposition to DRM must come from people indifferent to artists or intellectual property. His argument was instead that DRM was bad precisely because it attempted to protect creators and businesses by fighting the basic capabilities of computers and networks.
Twenty years later, Doctorow described the speech as a “defining document in the fight against DRM”. Its reputation is understandable. The talk is short, funny, technically accessible and startlingly confident. It also makes claims that can now be tested against two decades of evidence. Some were remarkably prescient; others were too absolute. The strongest modern reading is therefore more nuanced than Doctorow’s original slogan: DRM has proved poor at making inherently copyable digital information permanently uncopyable, but it has proved quite capable of controlling legitimate access, devices, markets and contractual relationships.
The Argument Doctorow Brings to Microsoft
Doctorow announces five propositions near the beginning and then builds the talk around them. DRM systems do not work, he says; they are bad for society, bad for business and bad for artists; and, finally, DRM is specifically a bad business decision for Microsoft. The order matters because each claim is meant to make the next one harder to dismiss.
The first section establishes the technical foundation. If DRM cannot securely prevent recipients from ultimately obtaining usable versions of digital content, Doctorow argues, then the restrictions it imposes must be judged by what they accomplish in practice rather than by the ideal of perfect copy protection. The second section asks who actually bears those restrictions and turns toward anticircumvention law, property and civil liberties. The third looks at interoperability and competition. The fourth asks whether DRM can nevertheless be justified as necessary for artists. The fifth converts everything into an argument aimed directly at Microsoft’s commercial interests.
That progression is the talk’s most important structural feature. Doctorow does not merely say that DRM inconveniences consumers. He tries to show that the technology begins from a defective security model, requires law to reinforce what technology cannot guarantee, gives rightsholders leverage over devices and competing businesses, misunderstands how new media create value, and finally asks companies such as Microsoft to weaken the general-purpose computers their own customers want.
The case is deliberately polemical. Doctorow is speaking to people inside a company with substantial investment in the technology he is attacking, so he uses jokes, provocations, stories and analogies rather than the careful qualifications of an academic paper. That makes the talk unusually memorable, but it also creates its most important weakness: several claims are phrased as absolutes when the evidence really supports narrower propositions.
Why Doctorow Says DRM Cannot Work
Doctorow begins far away from copyright. He starts with cryptography, because he wants the audience to see DRM not first as a moral or legal controversy but as a security problem.
Traditional cryptography can be simplified into three roles. Alice wants to send Bob a message, while Carol is the attacker who wants to read it. Early security systems may depend partly on keeping the cipher itself secret. Better cryptographic practice assumes that the attacker can know the algorithm while remaining unable to read the protected message without the key. The security burden is reduced to protecting something comparatively small and manageable rather than hiding the entire mechanism.
Doctorow emphasizes a principle associated with security expert Bruce Schneier: designers are bad judges of whether their own security systems are unbreakable. A system becomes trustworthy only after knowledgeable outsiders have been allowed to examine it, attack it and expose weaknesses. Security through obscurity is dangerous because a defender who does not know how the system has been broken may continue trusting it long after an adversary has compromised it.
DRM changes the conventional cryptographic situation in one crucial way. The intended recipient and the potential attacker become the same person.
Doctorow uses a DVD as the example. A studio can encrypt a film using the Content Scrambling System, or CSS, and sell the encrypted disc to a customer. But the customer does not want an encrypted collection of meaningless bits. The customer wants the movie. A DVD player therefore needs whatever machinery is required to turn the encrypted content into pictures and sound.
That requirement produces Doctorow’s central technical insight. In ordinary secure communication, Alice and Bob cooperate to hide information from Carol. In DRM, Alice gives Bob the encrypted file, gives Bob a device or program capable of processing it, and then asks the system to prevent Bob from obtaining unrestricted access to the information being rendered for him. The person against whom the secret is supposedly protected possesses the hardware on which the secret must eventually be used.
This does not mean every customer can effortlessly extract every protected work. Reverse engineering may require specialized knowledge, particular hardware, vulnerabilities, debugging tools or significant time. Doctorow’s claim is instead that the system contains an inherent endpoint problem: at some point, an authorized device must transform protected information into usable information.
Once one technically skilled person extracts a usable copy, the protection has failed in a second sense because digital networks allow that copy to spread. Doctorow therefore sees DRM as a system whose security repeatedly collapses at the place where encrypted content meets an authorized user.
That argument remains powerful, but his wording—“DRM systems don’t work”—bundles together several different meanings of success. A DRM system might fail to guarantee that no unrestricted copy will ever exist while still making extraction inconvenient for most customers. It might prevent ordinary copying inside authorized software, enforce rental expiration, restrict playback resolution, bind licences to devices or require a subscription to remain active. Whether that counts as “working” depends on the objective.
Modern DRM illustrates the distinction. Microsoft’s PlayReady architecture still encrypts protected content, but it also uses licence servers, client-specific licences, usage rules and controlled decryption. Microsoft can additionally employ revocation mechanisms for compromised clients and impose output restrictions on the paths through which protected video or audio leaves a device. These systems do not prove Doctorow wrong about the fundamental difficulty of giving someone readable information while guaranteeing that the same person can never capture it. They do show that commercial DRM evolved toward goals broader than permanent secrecy.
Doctorow’s technical section is therefore best understood as a strong argument against perfect copy control rather than proof that every form of DRM is economically or operationally useless. That distinction becomes increasingly important as the talk moves from cryptography into society and markets.
Why DRM Harms Users and Society
Doctorow anticipates the obvious response to his first argument: perhaps DRM does not need to stop experts. Perhaps it only needs to act as a “speed bump” that discourages ordinary people from casually copying protected media.
He rejects that defence because digital distribution separates the person who defeats the protection from the people who benefit from the defeat. An ordinary user does not need to reverse-engineer DRM if someone else has already produced an unrestricted copy and made it searchable online. One technically capable participant can therefore defeat a protection system on behalf of an enormous population of technically incapable users.
Doctorow points his Microsoft audience toward research produced by Microsoft employees themselves. Peter Biddle, Paul England, Marcus Peinado and Bryan Willman’s paper on “The Darknet and the Future of Content Distribution” examined networks through which users exchange digital material and argued that increases in bandwidth, storage, search and usability would make those networks progressively more effective. The implication for DRM was uncomfortable: once an unrestricted version entered a sufficiently capable sharing network, preventing every ordinary customer from personally breaking the DRM would not prevent the unrestricted work from circulating.
Doctorow then shifts from pirates to legitimate customers. One of his central stories concerns a mother who buys Toy Story on DVD and wants to make a VHS copy for her children so that the expensive DVD will not be destroyed through ordinary wear. She understands enough about electronics to connect a DVD player to a VCR, but Macrovision copy protection interferes with the recording.
For Doctorow, the story exposes a perverse asymmetry. The user who knows how to bypass the restriction can obtain circumvention hardware or an unrestricted copy. The person most effectively stopped by the DRM is the relatively unsophisticated customer who is trying to stay within an ordinary household use of something she bought. The protection is strongest precisely against the person least likely to be the industrial pirate the rhetoric of copy protection evokes.
Doctorow’s argument then becomes legal. If technological locks are routinely breakable, copyright industries can seek laws that prohibit people from breaking them. This is the role of anticircumvention rules, particularly Section 1201 of the U.S. Digital Millennium Copyright Act.
Section 1201 is more complicated than Doctorow’s broad rhetoric sometimes suggests, but the underlying legal mechanism is real. The U.S. Copyright Office’s current explanation of Section 1201 states that the law generally prohibits circumvention of technological measures controlling access to copyrighted works. A triennial process allows temporary exemptions for specified classes of noninfringing uses when the prohibition adversely affects them, so it is inaccurate to describe every act of circumvention as categorically forbidden under all circumstances. The important structural point, however, survives: the legality of breaking a technological access control can become a separate question from whether the underlying use of the work would itself infringe copyright.
Doctorow sees that separation as dangerous because software locks can acquire legal force. A rightsholder can design technical restrictions that prevent activities copyright law might not otherwise forbid and then invoke anticircumvention rules against tools that defeat those restrictions. In his account, code begins functioning like private legislation.
The Ed Felten episode gives the argument a scientific-research dimension. The Secure Digital Music Initiative invited researchers to attack proposed digital music protection technologies. Princeton professor Edward Felten and collaborators found weaknesses and prepared to publish their results. The researchers then received legal threats tied to the DMCA and withdrew their paper from one conference before pursuing legal protection for publication. The EFF’s complaint in the Felten case documents their contention that fear of litigation chilled scientific speech.
That story matters to Doctorow because security improves through adversarial scrutiny. If anticircumvention law creates incentives not to publish weaknesses in systems that protect copyrighted works, then a law intended to strengthen technological protection can interfere with the process by which security engineering discovers whether those protections are actually safe.
Dmitry Sklyarov’s case makes the point even more dramatically. Sklyarov, a Russian programmer, worked on software capable of removing restrictions from Adobe ebooks and was arrested in the United States after speaking at the DEF CON conference in 2001. An EFF account of the case and early DMCA enforcement records that he spent several weeks in jail and remained in the United States for months before being permitted to return to Russia. His employer, ElcomSoft, was later acquitted of the criminal charges brought against it.
Doctorow then widens the issue beyond books, films and music. Lexmark used technological measures in connection with toner cartridges and sued a company producing chips that enabled remanufactured cartridges to work with Lexmark printers. Chamberlain pursued a DMCA theory against Skylink over interoperable garage-door transmitters. To Doctorow, these cases demonstrated the danger of allowing copyright-related access controls to reach into markets for ordinary physical products.
The later outcomes are important because they complicate his warning without eliminating it. In Lexmark International v. Static Control Components, the Sixth Circuit vacated the preliminary injunction and rejected the idea that the DMCA could simply be transformed into a mechanism for monopolizing replacement parts. In Chamberlain Group v. Skylink Technologies, the Federal Circuit likewise rejected Chamberlain’s effort to establish DMCA liability over compatible garage-door controls. The cases therefore show both sides of the history: companies did attempt expansive anticircumvention theories, while courts sometimes imposed meaningful limits.
DVD region coding gives Doctorow a more intuitive example. A DVD purchased in one geographical region can be configured not to play in a machine assigned to another. Doctorow’s complaint is that the restriction is not equivalent to an ordinary copyright rule saying that consumers may never transport a legitimately acquired film across borders. It is a market segmentation policy enforced technologically and strengthened by legal controls surrounding circumvention.
The DeCSS controversy involving Norwegian teenager Jon Johansen extends the same point. Johansen helped develop software capable of decrypting CSS-protected DVDs. Doctorow presents the case as evidence of a world in which a person can become legally vulnerable for altering the relationship between a disc and a computer he owns. Johansen was ultimately acquitted in Norway, another reminder that Doctorow’s speech often captures the aggressive theory asserted by rightsholders more vividly than it captures the final legal boundary.
By the end of this section, the argument has moved well beyond piracy. Doctorow’s real concern is control over the owner’s machine. DRM, in his formulation, works only when a media company can decide what a customer’s computer, DVD player or other device will refuse to do. That is why he treats digital ownership as a social question rather than simply a licensing detail.
DRM, Interoperability and the Cost of Closed Markets
Doctorow’s third claim is that DRM is bad for business because innovation depends heavily on people being able to build new products that interact with existing products in unanticipated ways. He compares digital restrictions to historical efforts to control physical interfaces.
The telephone network supplies his clearest analogy. There was a period when customers could not simply plug arbitrary third-party devices into the telephone system. Once those restrictions weakened, an enormous market emerged for answering machines, cordless phones, novelty handsets, headsets and other equipment. The economic value came partly from allowing manufacturers to invent complements without obtaining permission from the company that controlled the original network.
Doctorow sees general-purpose computing as an extreme version of the same principle. The computer became extraordinarily valuable because software and hardware developers could make it perform tasks its original designers had never imagined. Open or broadly accessible interfaces allow one invention to become the platform on which another invention is built.
DRM introduces a conflicting principle. The company producing the recording—or film, ebook or other protected work—can acquire influence over the design of the “record player.” Devices must satisfy licensing requirements before they are permitted to render protected content, and those requirements can constrain copying, outputs, interoperability and supported features.
Doctorow finds it particularly strange that digital customers might evaluate a purchase partly by asking which proprietary ecosystem will continue authorizing the widest range of players. In a world of ordinary records or compact discs, a customer expects the medium to work across compatible equipment. Under DRM, compatibility becomes something granted or withdrawn through licences.
The DVD market becomes Doctorow’s example of what can be lost. DVD players became cheaper and smaller, but he argues that the tightly licensed ecosystem did not generate the same range of unexpected functionality that general-purpose computing produced elsewhere. Whether every example he offers proves a causal relationship is debatable, but the underlying economic question is important: how much complementary innovation disappears when innovators must obtain permission before building compatible products?
This is also where Doctorow’s argument is least reducible to “open is always better.” Closed ecosystems can create successful markets. Standardized restrictions can reassure rightsholders, simplify commercial licensing and produce consistent consumer experiences. The stronger claim is narrower: when the owner of one product can veto independent innovation around another product, some forms of experimentation become more expensive or disappear altogether.
Why Doctorow Thinks DRM Is Bad for Artists
Doctorow now confronts the argument with the greatest moral force behind DRM: even if restrictions are inconvenient, perhaps they are necessary to ensure that creators get paid.
He responds by telling a history of copyright repeatedly disrupted by new technology. His purpose is not to prove that creators never need protection. It is to show that protecting creators has often required changing the rules governing copyright rather than suppressing the technology that destabilized an older business model.
The player piano is his starting point. Piano rolls allowed musical performances to be mechanically reproduced on a large scale, threatening the sheet-music economy. Congress eventually responded not by granting composers an absolute veto over every mechanical reproduction but by establishing a compulsory licensing system. The Copyright Office’s history of the mechanical licence confirms that the 1909 Copyright Act created a compulsory licence after the first authorized mechanical reproduction and initially set a statutory royalty of two cents per copy.
That compromise matters enormously to Doctorow. The new technology could develop without requiring every manufacturer to negotiate a bespoke agreement for every composition, while composers still received compensation. Law adapted to the technical environment instead of forcing the technical environment to preserve the previous industry structure.
Doctorow sees versions of the same pattern in later media. Radio changed the economics of live and recorded music. Cable television redistributed broadcasts in ways the earlier system had not anticipated. New licensing arrangements and legal rules evolved because simply preserving every incumbent’s prior control would have prevented new industries from developing.
The VCR becomes his most famous example. Film studios feared home recording, and Jack Valenti of the Motion Picture Association notoriously compared the VCR’s threat to the film industry with the Boston Strangler. Yet the Supreme Court did not make Sony liable merely because Betamax recorders could be used to copy copyrighted programming.
Doctorow compresses the doctrine for rhetorical effect, so the precise legal holding matters. The Copyright Office’s account of the Sony Betamax decision explains that the Supreme Court held that secondary liability could not be imposed merely for supplying a device capable of substantial noninfringing uses. It also held that the television time-shifting at issue—recording a broadcast to watch it later—was fair use under the facts presented. The Court did not declare every possible form of home copying lawful.
The business history nevertheless supports Doctorow’s larger point. The VCR did not destroy filmed entertainment. Home video became an enormous commercial market, and the device that incumbent studios once regarded as existential eventually became another channel through which studios made money.
Doctorow uses these episodes to propose a theory of media change. Successful new media do not win because they reproduce every advantage of old media while eliminating every disadvantage. They win because they offer capabilities the old medium does not possess.
A printed book is extremely good at being a physical object: high resolution, portable without electricity, simple to lend, durable and easy to use. An ebook does not beat paper by becoming a slightly worse imitation of paper. Its distinctive advantages lie elsewhere: instant transmission, searchability, near-zero-cost duplication, storage, resizing, reformatting and the ability to move text through software.
The same comparison appears in music. MP3 sacrificed qualities people associated with records and CDs: packaging, liner notes, physical ownership and, depending on compression, some audio fidelity. But it provided something physical formats could not match—thousands of tracks in a pocket, easy copying between devices and transmission across networks.
Doctorow reduces this worldview to two facts about digital technology. Computers rearrange bits. Networks move bits between places cheaply and rapidly. A media system built on computers and networks therefore fights the medium when it tries to make bits difficult to manipulate or move.
This is the intellectual core of the talk. Doctorow believes that every mature medium eventually becomes organized around its native strengths. Attempts to make digital media imitate the scarcity of physical media may survive for a time, but the deeper economic pressure will favour businesses that exploit abundance, programmability and distribution.
His discussion of ebooks is striking in retrospect. In 2004, some arguments against widespread electronic reading emphasized display quality, the sensory pleasures of paper and attachment to physical libraries. Doctorow sees those concerns as analogous to complaining that radio lacks the social experience of a live concert. They may be true, but they describe what the older medium does well rather than what the new medium can uniquely offer.
Screen reading did, of course, become ordinary. Phones, tablets, laptops, dedicated ereaders and web browsers turned reading from screens into an everyday activity on a scale that would have been easy to underestimate in 2004. Doctorow’s broader intuition—that people would accept trade-offs in physical experience in exchange for the distinctive convenience of digital text—aged extremely well.
He then makes a stronger and more controversial leap. If copying and distribution become radically cheaper, Doctorow argues, culture can reach more people and support new creative opportunities. Technological disruption may damage businesses built around old inefficiencies, but society should not automatically preserve those inefficiencies by crippling the new technology.
This argument reflects his utilitarian conception of copyright. Copyright is not, in his telling, an eternal moral rule specifying the naturally correct price and permission structure for every use of creative work. It is a policy instrument intended to encourage creation and dissemination. If the technological environment changes, the best rules for producing that outcome may change with it.
The weakness lies in moving too quickly from “technology can produce a larger cultural market” to “artists will therefore be better off.” Those are not identical claims. Digital distribution can reduce costs and expand audiences while simultaneously shifting bargaining power toward platforms, aggregators or other intermediaries. A larger economic pie does not determine how the pie is divided.
Later media markets make that limitation difficult to ignore. The Internet unquestionably enabled enormous quantities of music, writing, video and other creative work to reach global audiences, but it also created new forms of platform dependence and concentration. Doctorow is strongest when he argues that old business models should not be treated as natural law. He is weaker when he implies that technological expansion by itself reliably solves the problem of creator compensation.
Still, the historical sequence does what he needs it to do. Piano rolls, broadcasting, cable and home video show that creative markets can survive transformations once described as existential threats. That gives him the foundation for his final move: if new technologies repeatedly create valuable new markets, Microsoft should not let entertainment companies persuade it to build deliberately constrained computers.
Why Microsoft Should Reject DRM
Doctorow’s final section turns directly toward the people in the room. His question is no longer whether DRM is philosophically justified. It is whether helping media companies control computers is actually a sensible strategy for a company whose business depends on selling computing platforms.
Sony becomes his cautionary example. Doctorow contrasts the Sony that introduced the Betamax VCR with the Sony that later owned major entertainment interests. The earlier hardware company was willing to build a machine consumers wanted despite Hollywood’s objections. The later Sony, in Doctorow’s account, became constrained by the interests of its music business and responded poorly to the rise of MP3 players.
He points to Sony digital music devices that supported proprietary or restricted formats when consumers increasingly wanted straightforward MP3 playback. His interpretation is that Sony spent engineering effort creating limitations no customer had requested. Protecting the entertainment division made the hardware less attractive.
Doctorow then aims the same criticism at Microsoft’s Windows Media ecosystem. A customer could rip or purchase music only to discover that hardware replacement, operating-system reinstallation or licence problems interfered with playback later. From the customer’s perspective, the system converted ordinary maintenance into uncertainty about whether legitimately acquired media would continue working.
His personal experience with Apple serves the same purpose. Doctorow describes reaching a limit on the number of computers authorized for his iTunes purchases because some of his machines had failed or were unavailable. The irony is central to his rhetoric: the restriction punished him partly because he was an enthusiastic Apple hardware buyer and an enthusiastic iTunes customer. Loyalty created more opportunities to collide with the licence boundary.
This leads to the product Doctorow actually wants: a player that plays everything.
The request sounds simple because that simplicity is part of his rhetorical strategy. A general-purpose computer is already technically capable of processing an extraordinary range of formats. DRM adds restrictions designed to ensure that some technically possible operations remain unavailable unless the relevant rights-holder approves them.
Doctorow argues that Microsoft should identify itself with its computer users rather than with entertainment companies seeking control over those users’ machines. Microsoft became powerful by selling general-purpose software and platforms capable of countless unexpected applications. Turning those machines into enforcement agents for third parties cuts against the source of their value.
The ringtone example pushes the logic to an almost absurd level. If someone owns an MP3 and a phone can technically play an MP3 as a ringtone, why should the phone be designed to prevent that use merely because a separate industry wants to sell ringtones? Doctorow sees such restrictions as business models masquerading as technical necessities.
His conclusion is consequently much more radical than a request for consumer-friendly DRM. He does not tell Microsoft to build a better digital lock. He tells Microsoft to stop asking permission and build the most capable player it can: the digital equivalent of a record player that plays everyone’s records.
The next two decades produced a mixed verdict. The market did reward extraordinary interoperability in some areas, especially general-purpose file formats and software. But Microsoft did not abandon DRM. Instead, it developed increasingly sophisticated infrastructure for protected commercial media, a fact that becomes central when the talk is reassessed from the present.
The Deeper Argument About Ownership, Competition and General-Purpose Computers
The five numbered claims ultimately depend on a deeper political-economic argument. Doctorow believes the central danger of DRM is not simply that a song becomes difficult to copy. It is that control over copyrighted content can become control over the machines, interfaces and markets through which people use that content.
That is why the speech repeatedly returns to record players, telephones, printers, garage-door openers and computers. Each example asks the same underlying question: when you buy one product, how much authority should the maker of another product retain over what your own property is permitted to do?
DRM as Control Beyond Copyright
Traditional copyright gives rightsholders specific legal powers, but those powers have limits. Purchasing a physical copy of a work does not normally mean purchasing the copyright itself, yet ownership of the physical object still carries practical freedoms. A reader can carry a book across borders, keep it for decades, resell that particular copy or use it without asking the publisher’s server for continuing authorization.
DRM can convert some of those practical freedoms into technical permissions. Whether a file can be opened, moved, printed, copied, played through a particular output or accessed after a certain date may be decided by software. The code can therefore prohibit conduct more broadly than ordinary users intuitively associate with copyright.
Doctorow’s concern becomes more serious when technological restrictions are combined with anticircumvention law. Section 1201 remains part of U.S. copyright law, although the Copyright Office’s triennial exemption process permits temporary exemptions for particular classes of noninfringing uses. The continuing need for that process illustrates the central issue Doctorow identified: a person may sometimes have a legitimate reason to use a work yet still confront a separate legal question because the desired use requires bypassing an access control.
The strongest version of Doctorow’s argument is therefore not that DRM literally creates new copyright statutes. It is that technical architecture changes the practical distribution of power. A publisher that could not realistically inspect how every printed book was used can design an ebook platform that automatically enforces rules over copying, exporting, printing or device access.
That distinction between legal rights and technical capabilities has only grown more important. Modern digital services frequently replace ownership-like transactions with licences and subscriptions. What appears to the customer as “my library” may depend on accounts, servers, device authorization and continued operation of a platform. DRM is one element of that larger shift from possession toward permission.
Interoperability as a Source of Competition
Doctorow’s business argument treats interoperability as an engine of innovation. When independent manufacturers can build something that connects to an existing platform, they do not have to persuade the platform owner that their idea deserves to exist before testing whether customers want it.
The printer and garage-door cases demonstrate the tension. Lexmark attempted to use copyright and DMCA theories against technology that allowed competing remanufactured cartridges to function with its printers. The Sixth Circuit’s decision in Lexmark International v. Static Control Components rejected the effort at that stage and explicitly warned against reading the DMCA so broadly that manufacturers could create replacement-parts monopolies by embedding copyrighted code behind technological locks.
Similarly, the Federal Circuit’s Chamberlain Group v. Skylink Technologies decision rejected Chamberlain’s DMCA claim involving a compatible universal garage-door transmitter. These cases matter because they reveal both the possibility Doctorow feared and a legal limit on it. Firms attempted to use access-control law to strengthen control over adjacent markets, but courts did not simply accept every theory.
The broader platform question remains. A closed interface can sometimes provide genuine benefits: security, reliability, consistent quality, privacy or straightforward support. But when permission is required for every complementary product, the platform owner gains the ability to decide not only whether something is safe but whether it competes with the platform owner’s own interests.
Doctorow therefore treats interoperability as a structural safeguard. A market in which the maker of records cannot dictate the design of every record player allows independent experimentation around the edges. Some experiments will be useless; others may create entire industries.
The point applies especially strongly to general-purpose computers because their value lies in programmability. A computer capable only of operations preapproved by every relevant media company is less general-purpose than one whose owner can install software, transform data and connect devices freely. Doctorow’s DRM argument is therefore also an argument about what kind of computer society wants.
Computers, Networks and the Economics of Abundance
Doctorow’s philosophy of digital media begins with material properties. Physical objects are rivalrous in ordinary use: if one person has a particular printed book, another person does not simultaneously possess the same physical copy. Digital information behaves differently because copying can reproduce the information without surrendering the original.
Networks magnify that difference. Once digital information can be duplicated cheaply and transmitted globally, distribution scarcity becomes difficult to preserve through ordinary physical limitations. DRM attempts to recreate some of that scarcity through software and licences.
Doctorow thinks businesses make a strategic mistake when they treat the properties of the new medium primarily as defects to suppress. A successful medium, in his view, eventually develops economic forms around what it makes easier.
That does not mean every digital product must be free. Cheap copying does not eliminate the costs of writing, recording, filmmaking, editing, marketing, hosting, software development or customer support. Nor does it eliminate people’s willingness to pay for reliability, convenience, curation, early access or participation in a service.
The more durable insight is that economics cannot pretend the copying cost is what it was in the physical era. Business models built around digital works must coexist with a technological environment in which reproduction and transmission are unusually easy. DRM can add friction, but it does not restore the underlying economics of a warehouse full of physical inventory.
This is one reason Doctorow’s argument still feels contemporary even though many of his examples are dated. The specific battle over DVDs and WMA has changed, but the conflict between technological capability and business-imposed restriction remains visible in subscription libraries, app stores, cloud services, game platforms, smart devices and digital media ecosystems.
How the Talk Works as Persuasion
Microsoft Research DRM Talk succeeds partly because Doctorow does not sound like someone presenting a policy memo. He sounds like someone trying to win an argument in a room full of technically sophisticated people who have heard the other side many times.
Humour lowers the defensive temperature. He greets the Microsoft audience as “fellow pirates,” jokes about his activist persona and populates his cryptography examples with deliberately absurd names. Later he mocks the idea that consumers might actively want companies to spend engineering resources making their own files harder to use.
The comedy is not decorative. It lets Doctorow frame certain industry assumptions as strange before he has fully argued against them. Once an audience laughs at the notion of evaluating music according to which company permits the widest choice of players, the burden of justification subtly shifts toward the DRM advocate.
His repeated “raise your hand” device works similarly. Doctorow imagines objections from the audience, invites people mentally to endorse them, then immediately responds. The speech becomes an argument with an invisible interlocutor rather than a monologue.
Analogies do much of the explanatory work. The record and record-player metaphor turns a complicated debate about licensing, cryptography and interoperability into a familiar relationship between media and devices. The telephone-jack example translates platform openness into everyday hardware. Piano rolls, VCRs and ringtones transform copyright policy into stories about technologies that audiences already understand.
This accessibility occasionally comes at the cost of precision. Doctorow prefers sharp binaries: DRM works or does not work; technologies are open or controlled; industries embrace new media or fight them. Real systems often occupy intermediate positions.
The Betamax discussion, for instance, gives the listener the correct broad lesson—that the law did not destroy the VCR to protect the incumbent movie business—but compresses a much narrower fair-use ruling. The legal cases surrounding the DMCA are presented mainly as evidence of danger, while later judicial limits receive less rhetorical attention because they were either unresolved in June 2004 or inconvenient to the polemical arc.
Anecdotes also carry substantial argumentative weight. The mother copying Toy Story and Doctorow’s own iTunes authorization problems make restrictions concrete, but they cannot by themselves establish how representative those experiences were. A more academic treatment would need stronger empirical evidence about consumer behaviour, infringement and market effects.
Yet academic caution would also destroy something valuable about the talk. Doctorow’s purpose was to force Microsoft employees to reconsider premises that had become embedded in products and policy. For that purpose, memorability mattered. The speech’s technical primer, historical narrative and jokes all serve the same persuasive design: make DRM appear not merely undesirable but conceptually upside down.
The result is unusually coherent. Doctorow starts by saying DRM asks a recipient’s computer to conceal information from its own owner and ends by asking Microsoft why it would want to sell computers whose competitive feature is obedience to somebody else. Even where the argument overreaches, the structure is exceptionally disciplined.
What Aged Well—and What Did Not
A fair retrospective should distinguish two questions. The first is whether Doctorow’s reasoning made sense given the technological and legal environment of 2004. The second is whether subsequent history confirmed the predictions embedded in that reasoning.
The answers are mixed in precisely the way that makes the talk still worth reading. Doctorow identified several durable structural problems with remarkable clarity, but he underestimated how effectively industries could adapt DRM from a promise of perfect copy prevention into infrastructure for controlling authorized access.
Did the Technical Argument Survive?
At the deepest level, yes. The endpoint problem remains. If a legitimate user can watch a film, hear a song or read an ebook, some authorized system eventually has to transform protected information into a perceptible or usable form. No DRM system has abolished that fact.
The history of unauthorized distribution also supports Doctorow’s belief that extracting one unrestricted copy can matter more than preventing millions of ordinary users from performing the extraction themselves. Modern piracy ecosystems do not require every participant to defeat the original protection. Copies can be ripped, captured, leaked or otherwise obtained by a smaller number of technically capable actors and then reproduced widely.
What did not survive is the implication that this makes DRM commercially pointless. Contemporary DRM systems pursue a portfolio of objectives rather than the single goal of making extraction impossible.
Microsoft’s PlayReady documentation describes a system capable of enforcing expiration rules, security levels and output restrictions. Its licence architecture can bind rights to clients or domains, while revocation can block compromised implementations from receiving new licences. These mechanisms may not guarantee that no pirate copy will ever exist, but that is not the only measure by which a streaming platform, film studio or device maker evaluates them.
A subscription service, for example, needs to distinguish current subscribers from former subscribers. It may want downloaded films to stop playing after a licence expires, high-resolution output to require a protected display path, or playback to occur only on authenticated clients. DRM can perform those functions even in a world where another person has uploaded the same film to an unauthorized network.
That means Doctorow’s technical argument survives best when restated carefully: client-side DRM cannot guarantee permanent control over information that must ultimately be rendered to the client. That is a powerful claim. It is not the same as saying DRM cannot enforce any useful restrictions.
Hardware-backed security also made the environment harder than the software-centric examples of 2004 suggested. Trusted execution environments, secure video paths, device certificates, hardware keys and server-side authorization can increase the skill and cost required to extract certain protected content. They do not repeal the endpoint problem, but they can change the economics of attacking it.
Doctorow therefore won the philosophical argument more decisively than the engineering-business argument. Digital information did not become permanently controllable simply because companies wanted it to be. Yet DRM vendors found commercially valuable territory between “perfect security” and “no control whatsoever.”
Did the Legal and Social Warning Survive?
Doctorow’s concern about anticircumvention law survived much more clearly. Section 1201 remains in force, and the United States still operates a recurring exemption process because legitimate activities can collide with the general prohibition against bypassing technological access controls.
The existence of exemptions is simultaneously evidence for and against the starkest version of his case. It demonstrates that the legal system recognizes noninfringing reasons to circumvent technological measures, but it also shows that the framework did not simply make every conceivable bypass permanently illegal.
The Felten and Sklyarov episodes remain important historical examples because they show that anticircumvention law can affect security research even when the underlying controversy is not ordinary commercial piracy. Doctorow’s warning that technological protection plus legal prohibition can chill research was not imaginary.
The competition cases are more nuanced. Lexmark and Chamberlain attempted interpretations that would have strengthened control over cartridges and garage-door transmitters, yet appellate courts rejected those particular theories. The legal system therefore proved more resistant to unlimited DRM-based aftermarket control than Doctorow’s rhetoric sometimes implies.
Still, his underlying concern has widened beyond the media controversies of 2004. Software now sits inside cars, appliances, agricultural equipment, medical devices and countless consumer products. Questions about repair, interoperability, access controls and owners’ ability to modify software-enabled possessions are no longer confined to DVDs or ebooks.
The most durable social insight is that technological enforcement changes practical power even before a lawsuit occurs. If a device refuses an operation automatically, most users will never litigate whether they would have had a legal right to perform it. Code can therefore shape behaviour more immediately than formal legal rules.
That is why Doctorow’s distinction between copyright and control remains valuable. Copyright may authorize particular exclusive rights, while technical architecture determines what users can actually do. The two overlap, but they are not identical.
Did Doctorow Get the Media Business Right?
This is where the record is most interesting because different media moved in different directions.
Purchased digital music strongly vindicated Doctorow. In January 2009, Apple announced that the iTunes Store would make its music catalogue available without DRM. A market that had once been built around tightly controlled downloads shifted toward ordinary audio files that could be copied among compatible devices much more freely.
Doctorow’s complaints about device authorization and restricted music files consequently look prescient. Consumers wanted music to survive hardware changes and play across devices, and the purchased-download market eventually moved much closer to that expectation.
His broader prediction about reading from screens also aged extremely well. The idea that people would reject digital text because paper was nicer underestimated the value of carrying enormous libraries on small devices, downloading books instantly, searching text and changing font size. Electronic reading became normal even though paper retained many of its advantages.
But DRM did not disappear from ebooks. Amazon’s current Kindle publishing documentation allows publishers to choose DRM while also giving DRM-free books greater portability; from January 20, 2026, verified purchasers of confirmed DRM-free Kindle books can download EPUB and PDF files. The market therefore developed into a mixed ecosystem rather than a simple victory of unrestricted files over protected ones.
Video diverged even more sharply from Doctorow’s expected trajectory. Streaming became one of the dominant ways people consume films and television, and streaming services commonly use DRM because the commercial transaction is not ownership of a permanent unrestricted file. The customer receives conditional access to a catalogue.
That business model changes the consumer expectation. If someone pays for one month of a streaming service, few customers believe they have purchased an irrevocable personal copy of every film they watched during that month. DRM fits subscription access more naturally than it fit the rhetoric of “buying” a song that could later become inaccessible because a computer authorization failed.
Microsoft’s own history is the clearest contradiction to Doctorow’s specific strategic recommendation. The company did not walk away from DRM. PlayReady remains an actively documented and licensed technology, with modern rules governing licences, clients, revocation and output protection. Microsoft evidently found continuing business value in providing the infrastructure that Doctorow urged it to reject.
Yet Microsoft also continued selling general-purpose computing platforms. The outcome was not a clean choice between universal computers and DRM machines. Modern devices often contain both: enormously flexible computing environments alongside protected pathways designed to restrict particular classes of commercial media.
This coexistence exposes one of the assumptions in the talk. Doctorow tends to imagine that DRM must eventually lose because consumers prefer more capable devices. Consumers do generally prefer capability, but they also value convenience, exclusive content and integrated services. A restrictive ecosystem can prosper if the surrounding product is attractive enough.
The success of closed or partially closed platforms therefore weakens any simple theory that openness automatically wins markets. Compatibility matters, but so do network effects, catalogues, user experience, brand loyalty, security, pricing and distribution.
Doctorow was more accurate about the direction of digital media than about the disappearance of technical control. Media became more networked, more screen-based, more instantly distributable and less dependent on physical containers, exactly as he expected. But the business response was often not to surrender control over digital files. It was to change the transaction from selling files to selling authenticated access.
That is perhaps the biggest development the 2004 speech does not fully anticipate. DRM proved especially durable once media companies stopped promising customers something closely resembling ownership.
Critical Review: How Strong Is the Microsoft Research DRM Talk Today?
As a piece of explanatory writing, Microsoft Research DRM Talk is exceptional. Doctorow takes an intimidating mixture of cryptography, copyright law, technological history and platform economics and turns it into a single argument that a non-specialist can follow. The recipient-as-attacker model is especially effective because it gives readers a conceptual tool rather than merely a conclusion.
Its organization is equally strong. Each section answers a different possible defence of DRM. Technical effectiveness leads to social consequences; social consequences lead to business effects; business effects lead to artists; the artists argument finally leads back to Microsoft. The progression prevents the speech from becoming a loose collection of complaints.
Doctorow is also remarkably good at choosing examples that make abstract systems visible. DVD region coding shows the difference between copyright and technologically enforced market segmentation. The Toy Story copy demonstrates how restrictions can fall hardest on unsophisticated legitimate users. Printers and garage-door openers demonstrate why anticircumvention matters beyond entertainment.
His media analysis is one of the strongest parts of the work. The insistence that new media should be understood through the capabilities they add rather than the qualities they lose remains a useful framework. His dismissal of the idea that people would resist electronic reading merely because physical books were more pleasant now looks especially perceptive.
The speech also deserves credit for treating copyright historically rather than as a fixed natural order. Piano rolls, broadcasting and home video really did force institutions to adapt. Doctorow is right that industries frequently describe disruptive technologies as threats before discovering ways to profit from them.
His strongest contribution, however, is the connection between DRM and general-purpose computing. That insight has only become more important as software has spread into ordinary objects. A technological protection system is never solely about the protected song, film or program; it can also determine what the surrounding machine is permitted to do.
The most important flaw is the phrase around which the entire talk is organized: DRM “doesn’t work.” It is rhetorically powerful but analytically underspecified.
If “work” means permanently preventing any recipient from extracting an unrestricted copy, Doctorow’s argument is formidable. If it means enforcing rules for most legitimate clients, controlling subscription access, limiting outputs, expiring rentals, binding licences to devices or satisfying contractual requirements imposed by content providers, modern systems clearly can work in meaningful ways.
The speech would be stronger if it distinguished those goals explicitly. DRM can fail as an anti-piracy absolute while succeeding as a mechanism for market governance. Indeed, that may be the more troubling interpretation of Doctorow’s own argument: the technology’s greatest power may never have been its ability to eliminate unauthorized copies, but its ability to define what authorized customers are permitted to do.
The historical argument also selects its evidence strategically. Technologies such as the player piano and VCR ultimately produced large new markets, but not every disruption creates a prosperous replacement system for creators. The fact that old intermediaries lose control does not guarantee that artists receive a larger share of the resulting value.
That omission matters because Doctorow repeatedly suggests that bigger technological markets mean a bigger pie for artists. The first part is often true; the second depends on bargaining power, contracts, platform concentration and how revenue is distributed. Digital abundance can create millions of potential listeners while still leaving many creators economically precarious.
His treatment of openness has a similar tendency toward overgeneralization. Permissionless interoperability can produce extraordinary innovation, but closed systems sometimes produce real benefits, including security, consistency and ease of use. The difficult policy question is not whether every restriction is inherently bad but when the benefits of control justify the innovation and autonomy sacrificed to obtain it.
The legal material needs comparable qualification. Felten and Sklyarov show genuine dangers surrounding anticircumvention, but cases such as Lexmark and Chamberlain also demonstrate that courts did not accept every attempt to turn copyright law into control over adjacent hardware markets. The legal history is therefore less uniformly bleak than the speech’s polemical momentum suggests.
None of these weaknesses destroys Doctorow’s central project. They make the argument less absolute, but they actually sharpen its most valuable insight.
The lasting question is not simply whether DRM can be cracked. Almost every important technology policy debate eventually becomes more complicated than that. The lasting question is who gets to decide what a programmable device may do after the device reaches its owner.
That question remains visible everywhere. A computer can technically copy, modify, convert, transmit and store information. A platform can technically allow third-party software. A device owner may technically be capable of repairing or modifying a product. Whether those capabilities remain available depends increasingly on code, licences, servers, contracts and regulation.
Doctorow understood unusually early that digital copyright disputes were becoming disputes about control over computers themselves. That is why the talk remains more important than many of its obsolete products and references would suggest.
It is also why the sections that aged poorly do not make the work obsolete. The persistence of DRM does not refute the entire speech; it forces a better interpretation of it. Companies did not solve the basic technical contradiction Doctorow described. They learned to build commercially useful systems around it.
The fairest overall judgment is therefore that Microsoft Research DRM Talk is both more right and less right than its most famous claim suggests. Doctorow was too categorical about the commercial uselessness of DRM, too optimistic that open systems would consistently defeat controlled ecosystems, and too quick to assume that technological abundance would translate into prosperity for artists.
But he was remarkably prescient about several deeper trends. People did embrace screen reading. Purchased digital music moved away from DRM. Copy protection did not make unrestricted copies disappear. Anticircumvention remained controversial. Interoperability became a central policy issue far beyond entertainment media. Software increasingly became a way for manufacturers and platforms to determine what owners could do with their own devices.
Most importantly, he recognized that DRM was never merely a battle between pirates and creators. It was a battle over whether the programmable machines of the digital age would primarily obey their owners or the remote parties whose content and services happened to pass through them.
For readers interested in digital ownership, copyright, platform power, interoperability or the history of technology policy, the talk remains worth reading. Its examples belong unmistakably to the era of DVDs, WMA files, early ebooks and the original iTunes Store, but the central conflict has not disappeared. It has become embedded more deeply in everyday computing.
Doctorow ended by asking Microsoft to build the digital equivalent of a record player that could play everyone’s records. The industry largely chose a more complicated future: extraordinarily capable general-purpose machines containing increasingly sophisticated systems for deciding when those capabilities must be withheld. That outcome does not vindicate every part of his argument, but it makes the question behind the argument harder to dismiss than ever.
Last Updated on October 5, 2026 by Aseem Gupta