The History of Sex Dolls: From Rubber Women to RealDoll, Robots, and AI

Answer first

Five thresholds, not one inventor.

“First” changes when the question moves from imagined partner to documented commodity, mass market, premium platform, or interactive machine.

1. 01Artificial partner Myth, literature, and automata—not product evidence
2. 02Commercial class Late-19th-century French inflatable rubber women
3. 03Mass market Postwar plastics, mail order, novelty, and adult retail
4. 04Premium platform Japanese love dolls, silicone, skeletons, and customization
5. 05Interactive system Prototype, delivery, autonomy, and adoption remain separate tests

Best-supported origin answer: the earliest securely documented commercial class in the present source base is the French femme en caoutchouc, not an ancient statue, a sailor-made doll, or the alleged Borghild Project.

Direct answer: who invented the sex doll, and when? #

This is one device-family history within the broader history of sex toys. No single person invented the sex doll. The answer changes according to what is being counted. Ancient and early-modern literature imagined artificial partners, but a story such as Ovid’s Pygmalion is not evidence that people manufactured a sexual device. Private human simulacra may once have been made from textile, leather, wood, wax, or stuffing, yet those materials survive poorly and famous claims about sailors’ homemade dames de voyage lack secure contemporaneous documentation. The earliest commercial class that Bo Ruberg’s archival study identifies in this source base is the inflatable rubber woman sold in late-nineteenth-century France. Bo Ruberg’s archival history treats these femmes en caoutchouc as documented commodities and shows why the better-known sailor-origin story should be classified as an imagined history rather than a settled fact. STRONGLY SUPPORTED. [S001] [S037]

Several later objects are historically important without being “the first.” Oskar Kokoschka commissioned a life-size figure modeled on Alma Mahler in 1918–1919; it belonged to private, artistic, tactile, and performative history rather than mass-market manufacturing. Hans Bellmer’s articulated dolls of the 1930s were eroticized Surrealist artworks, not ordinary retail sex dolls. Cheap postwar plastic inflatables widened access, Japanese specialist manufacturers developed the modern “love doll” market from the late 1970s, and Abyss Creations’ RealDoll became an influential premium silicone-and-skeleton platform in the mid-1990s. [S019] [S022] [S023] [S056]

The first sex robot is also unresolved. A static doll, a doll with prerecorded speech, a remote-controlled mechanism, a sensor-responsive machine, and an embodied conversational system are different technical thresholds. Roxxxy was publicly introduced in 2010 as a sex-robot prototype, but announcement, autonomy, delivery, and durable commercialization are not interchangeable. Harmony and later Realbotix systems brought together conversational software and robotic components, but present capabilities and availability require publication-date verification. [S004] [S058] [S060] [S061]

The defensible conclusion is therefore plural: artificial partners were imagined long before a modern industry existed; purpose-built commercial rubber dolls are securely documented by the late nineteenth century; mass-market plastic dolls expanded after the Second World War; articulated silicone dolls became prominent in the 1990s; and interactive robotic or AI-linked systems remain a developing, technically contested category.

Executive thesis #

The history of the sex doll is not a straight line from an ancient statue to an intelligent robot. It is a history of category formation. Different societies and industries produced statues, mannequins, anatomical models, automata, artist’s dolls, inflatable commodities, novelty products, companion objects, lifelike silicone bodies, and interactive machines. Only some were designed or used primarily for sexual purposes. Human shape alone does not establish sexual function.

Two propositions survive the evidence. First, the modern sex doll emerged not from one invention but from the convergence of artificial-person fantasies, rubber and plastics technology, molding and inflation, obscenity-controlled commerce, specialist retail, changing sexual cultures, modular manufacture, and later robotics and software. Second, origin stories have done political work of their own. Tales about lonely sailors, national inventors, military programs, or solitary heterosexual men have often displaced more complicated histories involving sex workers, artisans, manufacturers, queer and disabled users, colonial language, commercial secrecy, and companionate use. [S001] [S003] [S004]

A source-critical history must therefore separate five questions:

1. When did people imagine artificial partners?
2. When do physical human simulacra appear in reliable evidence?
3. When can an object be shown to have a sexual purpose?
4. When did purpose-built dolls become commercial products?
5. When did dolls become interactive robotic or software-connected systems?

Collapsing those questions creates almost every familiar “first sex doll” myth.

What is a sex doll? #

For this history, a sex doll is a three-dimensional human or humanoid object designed, marketed, or demonstrably used primarily for sexual activity. The definition has three safeguards.

First, design is not inferred from body shape alone. A mannequin, anatomical model, effigy, fashion doll, sacred image, puppet, artist’s doll, or funeral figure can be anatomically detailed without being a sexual implement. Second, use must be supported by context: advertising, instructions, a reliable testimony, a court or customs record, a patent, a documented commission, or design features that become persuasive when combined with other evidence. Third, “primarily” allows secondary histories. A display mannequin can be used sexually by an individual, but that does not retroactively make the whole mannequin category a sex-toy invention.

An inflatable sex doll creates its volume through air pressure inside a sealed flexible skin. A love doll is a later commercial and cultural label, especially significant in Japan, whose meanings may include sexual use, photography, collecting, dress, care, and companionship. A sex robot adds powered actuation and some degree of sensing or programmed interaction; a motionless doll with a speaker is not necessarily a robot. An AI companion is primarily software. It belongs within sex-doll history only when it is integrated with or materially organized around an embodied doll or robot platform. [S004] [S011] [S023]

The term artificial partner is deliberately broader. It can include myths, fictional androids, dolls used for companionship, and embodied machines without assuming sexual function. That broader category is useful for cultural history but should not be mistaken for a product class.

How historians know—and why the evidence is uneven #

Sex-doll history has an unusually difficult archive. Products were private, stigmatized, easily destroyed, and often made from perishable materials. Cloth decomposes. Rubber oxidizes, cracks, and becomes tacky. Plastic films split at seams. Owners may discard objects without documentation. Retailers used euphemism to avoid police, postal restrictions, payment barriers, or embarrassment. Museums historically collected elite art, medical instruments, and industrial design more readily than adult commodities.

Different source types answer different questions:

• Objects can establish material, dimensions, joints, repair, and manufacturing, but function may remain uncertain.
• Advertisements and catalogs prove that a seller offered an item and reveal price, language, audience, and distribution. They do not prove how many were sold or how buyers used them.
• Sexological and medical texts preserve descriptions otherwise lost, but they may mix observation, moral judgment, anecdote, translation, and hearsay.
• Patents show a claimed design and legal filing. They do not prove manufacture, sale, technical success, or priority over unpatented practice.
• Art and literature reveal fantasies, anxieties, conventions, and symbolic meanings. They cannot be treated as documentary photographs of ordinary behavior.
• Court and customs files document prohibited or contested trade, but overrepresent what authorities targeted.
• Company archives clarify product lines and brand chronology while also serving marketing interests.
• Interviews and owner studies reveal experience, but samples are often small, self-selected, and culturally specific. [S004] [S037] [S103] [S104]

The best conclusions rest on converging evidence. A dated advertisement plus a physical object plus a customs record is stronger than a later anecdote. A modern blog citing another modern blog is not an origin source.

Artificial people before sex dolls #

Pygmalion: a myth, not an artifact report #

In Book 10 of Ovid’s Metamorphoses, Pygmalion makes an ivory female figure, falls in love with it, and sees it brought to life by divine action. The story became one of the most influential Western narratives about desire, artistry, animation, and the idealized artificial woman. Later painters, writers, dramatists, filmmakers, roboticists, and journalists repeatedly used “Pygmalion” as shorthand for an artificial partner. [S033] [S014]

Calling the ivory figure “the first sex doll,” however, imports a modern object category into literary myth. The narrative does not document a surviving device, workshop practice, commercial market, or archaeological class. It concerns transformation from representation into living personhood. Its historical value lies in the long cultural genealogy of artificial-partner desire, not in proving ancient manufacture.

The same distinction applies to Hephaestus’s golden attendants in the Iliad. They are imagined artificial servants endowed with intelligence and motion, important to the history of automata and artificial life, but not sexual devices. [S034] [S015]

Claim classification: Pygmalion as an artificial-partner myth is DOCUMENTED. Pygmalion as evidence for an ancient sex doll is UNSUPPORTED.

Automata, mannequins, and anatomical bodies #

European and Asian societies built moving figures, devotional images, anatomical models, puppets, waxworks, and display mannequins long before modern adult retail. Enlightenment androids performed music, wrote, drew, or simulated bodily processes. These machines mattered because they made human likeness, motion, and mechanism commercially and philosophically visible. They also shaped later fiction about artificial women. [S015] [S016] [S017]

Yet an automaton becomes relevant to sex-doll history only when erotic purpose is documented. The frequent modern move from “human-shaped machine” to “sex robot ancestor” confuses technological resemblance with historical function. A chess-playing automaton and a sex robot both simulate agency, but they belong to different institutions, markets, and use histories.

E. T. A. Hoffmann’s Olympia in “The Sandman” and Villiers de l’Isle-Adam’s artificial woman in L’Ève future intensified nineteenth-century anxieties about deceptive femininity, idealization, technology, and male control. They are indispensable to the cultural history of artificial partners, while remaining fictional. [S035] [S036] [S014]

The sailor-made dames de voyage story #

The most repeated origin account says that French, Spanish, Dutch, or other European sailors made crude artificial women from cloth, leather, straw, or old clothing during long voyages. These objects are commonly called dames de voyage—“ladies of the voyage.” Retellings often supply a century, nationality, or material without identifying a logbook, diary, ship inventory, court record, surviving object, or contemporaneous illustration.

Ruberg’s Sex Dolls at Sea subjects this tradition to source criticism and argues that the sailor narrative is an imagined history: a story through which later writers explained sexual technology by projecting it onto isolated men at sea. The narrative connects artificial women to maritime masculinity, colonial movement, racialized assumptions, and the fantasy that sex technology arises naturally where women are absent. [S001]

That does not prove no sailor ever improvised a sexual object. Private, perishable objects often leave no archive, and categorical denial would exceed the evidence. It does mean that “sailors invented the sex doll” cannot responsibly be published as fact. The correct distinction is:

• POSSIBLE: individual improvised simulacra may have existed aboard ships or elsewhere.
• INSUFFICIENT EVIDENCE: no securely established early object or primary record identifies a first sailor-made doll.
• UNSUPPORTED: assigning invention to Dutch or French sailors on the basis of later repetition.

The methodological lesson extends beyond this topic. Plausibility is not provenance. A practice may be imaginable without being historically demonstrated.

“Dutch wife”: language is not an invention record #

“Dutch wife” has been used for objects other than sex dolls, including elongated bed supports and bamboo or wicker devices intended to keep a sleeper cool or separate limbs in hot climates. Later usage acquired sexual and colonial associations, and related expressions circulated in European and Asian languages. Historical dictionaries document changing meanings, but dictionary evidence must be read alongside colonial context and object history. [S106] [S107]

Two errors are common. The first treats “Dutch” as evidence that Dutch people invented a sexual doll. The second assumes every object called a Dutch wife was a human simulacrum. Neither follows. Ethnic adjectives in English often emerged through rivalry, stereotype, or derogatory comparison rather than accurate attribution. A cooling bolster is not a sex doll merely because later speakers reused or sexualized the phrase.

The recommended public wording is cautious: “Dutch wife” is a historically layered expression applied to bed supports and, in later contexts, artificial sexual partners. Its etymology does not identify a single inventor or nation of origin. The term should also be flagged as colonial and potentially derogatory rather than presented as neutral technical vocabulary.

Rubber and the first securely documented commercial dolls #

What vulcanization changed #

Charles Goodyear’s 1844 U.S. patent forms part of the history of vulcanized rubber, a process family that made natural rubber more stable across temperature ranges and more useful for durable industrial goods. Rubber history also includes plantation labor, imperial extraction, chemical compounding, patents, and large manufacturing systems; it should not be reduced to an inventor anecdote. [S027] [S028] [S029] [S049] [S050]

For artificial bodies, rubber offered several advantages over stuffed textile or rigid carving. It could be made flexible, molded, joined into sealed envelopes, inflated, colored, washed, packed, and repaired. These possibilities did not automatically produce a sex doll in 1844. Enabling technology and product invention are separate stages.

Femmes en caoutchouc #

The strongest currently available origin point for a commercial purpose-built sex doll is the late-nineteenth-century French market for inflatable rubber women. Ruberg reconstructs these commodities through archival and printed evidence, using the phrase femmes en caoutchouc—rubber women—and distinguishes them from the unproven shipboard narrative. Iwan Bloch’s early-twentieth-century sexological writing also describes artificial substitutes and pneumatic products, though his text must be read as a period source rather than an objective market survey. [S001] [S037]

These products matter for three reasons. First, they were made for sale, not merely imagined. Second, inflation allowed a human-scale body to be shipped and stored in relatively compact form. Third, commercial evidence connects material, purpose, seller, and audience more securely than shape alone.

The archive still does not give a neat inventor, first factory, or universally dated launch. Adult goods often moved through euphemistic, semi-clandestine, and poorly preserved networks. A responsible formulation is therefore:

The earliest securely documented commercial sex dolls identified by Ruberg’s archival study are inflatable rubber women sold in late-nineteenth-century France. This is not proof that no earlier private object existed.

That answer is more useful than a sensational “first” because it states the threshold and the limits.

Iwan Bloch and the sexological archive #

Iwan Bloch’s The Sexual Life of Our Time in Its Relations to Modern Civilization appeared in German in the early twentieth century and in English translation in 1909. Like other sexological texts, it attempted to classify sexual behavior and devices within a medical, social, and moral system. Its descriptions of artificial substitutes are valuable because products made of rubber and fabric rarely survive. [S037]

Bloch did not invent the sex doll. Nor should every reported object in a sexological compendium be accepted without qualification. Sexologists relied on case histories, correspondence, police material, commercial advertising, literature, and reports from other writers. Their categories were shaped by period assumptions about normality, pathology, gender, race, and prostitution.

The correct use of Bloch is evidentiary triangulation. When his account aligns with advertisements, catalogs, or other contemporary reports, it strengthens the case that an adult commodity existed. When a dramatic anecdote appears only in a later translation or unsupported retelling, it should be quarantined. Sexology supplies vocabulary and observation, not automatic certainty.

Oskar Kokoschka and the Alma doll #

After his relationship with Alma Mahler ended, the Austrian artist Oskar Kokoschka commissioned the doll-maker Hermine Moos to construct a life-size female figure resembling Mahler. His letters supplied unusually detailed instructions concerning body form, skin-like surfaces, hair, touch, and clothing. The figure was delivered in 1919, appeared in photographs and paintings, and was incorporated into private and performative episodes before being destroyed. [S022] [S041]

This is one of the best-documented singular artificial bodies in modern cultural history. It is also resistant to one-label summaries. The commission involved grief, control, substitution, artistic experimentation, tactile desire, performance, and self-mythology. Scholars have interpreted it through gender, authorship, fetishism, modernism, and the relation between male artist and female maker. [S022]

Calling it simply “the first modern sex doll” loses most of what the evidence shows. Commercial inflatable rubber women had already been described. Moos’s object was bespoke rather than standardized, and its surviving documentary life is artistic. The phrase “sex toy” may describe one dimension of Kokoschka’s intention, but it should not erase the object’s status as a commissioned artwork and simulacrum.

Classification: DOCUMENTED private artificial body with erotic and tactile dimensions; not a first commercial sex doll.

The case also exposes a recurring bias: histories often center the famous male commissioner and neglect the woman who fabricated the object. Hermine Moos’s labor, materials knowledge, and translation of contradictory instructions were essential to the doll’s existence.

Hans Bellmer’s dolls #

Hans Bellmer began constructing and photographing articulated dolls in the early 1930s. His works disassembled and recombined the female body through ball joints, duplicated torsos, limbs, and staged photographs. They became central to Surrealist art and have been interpreted in relation to desire, violence, fascism, the uncanny, childhood, fragmentation, and power. [S019] [S020] [S021] [S042] [S043]

Bellmer’s dolls are directly eroticized, but erotic art is not automatically a functional sex device. Their provenance lies in studio production, photography, publication, exhibition, and art history. No responsible product chronology should place them between an inflatable retail doll and RealDoll as though all three were equivalent consumer goods.

This distinction does not sanitize the works. Their sexual and violent meanings are central to interpretation. It simply keeps representation, art object, and sexual implement analytically separate. The correct label is documented eroticized art doll, with practical sexual use unproven and unnecessary to its historical importance.

The Borghild Project: a citation autopsy #

A widely circulated story claims that Nazi Germany created a secret “Borghild Project” around 1941 to supply Aryan-looking inflatable dolls to soldiers, prevent sexually transmitted infections, or discourage contact with non-German women. Versions add an approving Heinrich Himmler, a Danish or German designer, precise dimensions, field tests, and destruction of a factory in an Allied bombing raid.

The narrative has the structure of documentary history but lacks the evidentiary architecture expected for a military procurement project. Credible scholarship has not produced a contemporaneous order, budget, specification, patent, manufacturer contract, personnel file, test report, photograph, surviving object, or archive citation. The story circulated through internet pages and popular retellings whose details expanded as they were copied. Ruberg and Lieberman treat such origin narratives critically rather than accepting them as wartime fact. [S001] [S003]

The proper classification is UNSUPPORTED / likely modern hoax. “Likely” preserves the logical possibility that unknown material could surface, but the burden rests on the claimant. A source that merely repeats the story is evidence of circulation, not corroboration.

Recommended wording:

No reliable contemporaneous evidence has established the alleged Nazi Borghild Project. It should not be included in a historical timeline except as a modern origin myth and example of citation laundering.

This claim illustrates why precision matters. A false story can dominate search results because its details are memorable, while the actual late-nineteenth-century rubber market is harder to summarize.

Bild Lilli was not a Nazi sex doll #

Bild Lilli began as a character in Reinhard Beuthien’s comic for the West German newspaper Bild and became a three-dimensional fashion doll in the 1950s. Museum object records document its West German manufacture, 1955 production date, and place in Barbie’s design lineage. [S045] [S046] [S047] [S048]

The phrase “adult novelty” is sometimes converted into “sex doll,” and the doll’s German origin is then distorted into a Nazi connection. The chronology alone defeats the claim: Lilli was a postwar West German product. Her stylized body and flirtatious comic persona can be analyzed as sexualized consumer culture, but she was not a human-scale implement designed for sexual use.

Classification: fashion/novelty doll with adult cultural associations; not a functional sex doll and not a Nazi product.

Postwar plastics and the inflatable mass market #

The nineteenth-century rubber woman was a comparatively expensive and technically demanding commodity. Postwar plastics transformed the economics of inflation. Plasticized vinyl films could be produced in thin sheets, printed, cut, and joined with heat or radio-frequency welding. A life-size hollow form could be folded into a small package, shipped cheaply, and sold through mail order, novelty outlets, adult bookstores, and men’s-magazine advertising.

The surviving record is fragmented. Products were disposable, sellers changed names, catalogs were discarded, and patents often described generic inflatable dolls without explicit sexual claims. That makes a single “first vinyl sex doll” difficult to establish. Patent search results prove design filings, not market success, and many makers relied on trade secrecy or ordinary packaging technology rather than patent protection. [S003] [S064] [S065] [S108]

What can be said with confidence is that postwar material and retail systems changed scale. Printed faces and bodies substituted graphic suggestion for sculpted anatomy. Low price encouraged gag-gift marketing alongside sexual use. The doll could be simultaneously presented as obscene merchandise, comic novelty, bachelor accessory, or private aid. Packaging helped manage that ambiguity.

This phase produced the “blow-up doll” that dominates late-twentieth-century comedy. Its cultural visibility should not be mistaken for the origin of the sex doll. It was a mass-market redesign of an older commercial principle: the inflatable artificial body.

Mail order, adult retail, and euphemism #

The separate history of specialist and feminist sex shops shows how retail also became an information system built around education, curation, and customer knowledge.

Adult commodities developed under legal and logistical pressure. Obscenity rules, postal restrictions, customs enforcement, zoning, payment barriers, and social stigma encouraged sellers to describe products as novelties, companions, models, marital aids, or gifts. Euphemism was not merely coy language; it was part of the distribution technology.

Mail order gave buyers distance from a face-to-face transaction and allowed sellers to reach places without specialist shops. It also created risks: packages could be intercepted, catalogs prosecuted, customer lists exposed, and product claims exaggerated. Adult bookstores later provided a physical retail channel but often organized products around a heterosexual male customer imagined by the proprietor or publisher. Feminist and queer sex-toy stores challenged that model by adding education, material information, and different aesthetic expectations, although dolls were not always central to their product mix. [S031] [S032]

The same object could acquire different meanings depending on channel. A rubber figure in a medical supplier’s catalog, a novelty advertisement, a pornographic shop, an art gallery, and a disability-focused design discussion did not enter the same social world. Retail location classified the object for both law and buyers.

Japan and the modern “love doll” #

Japan became a major center of specialist life-size doll design and cultural discussion. The English phrase “sex doll” does not map perfectly onto Japanese terms and histories. Older expressions related to the “Dutch wife” circulated before manufacturers and users increasingly adopted love doll (rabu dōru). Agnès Giard’s research shows that modern Japanese dolls can be embedded in practices of dress, photography, naming, care, memorialization, fantasy, and companionship as well as sexual use. [S023]

Orient Industry traces its business to 1977 and became one of the best-known Japanese manufacturers. Company histories are useful for models and dates but should be corroborated because brands construct their own heritage. [S059] [S023]

Japanese development also complicates a Western progression from crude inflatable to “realistic” silicone body. Design choices reflect aesthetics, pornography, animation, photography, domestic space, storage, weight, maintenance, and imagined personality. A doll’s meaning is not reducible to anatomical accuracy. Some owners emphasize a character or presence; some treat the object as a photographic model; some use it sexually; practices can overlap.

Media accounts often present Japanese love dolls as evidence of national loneliness or demographic decline. That framing converts a heterogeneous product culture into a story about an allegedly strange society. A responsible history uses Japanese scholarship and owner testimony, avoids national diagnosis, and distinguishes documented practices from journalistic symbolism.

From latex and foam to silicone and articulated bodies #

Late-twentieth-century manufacturers experimented with latex, foam, vinyl, and composite constructions to create bodies that were more substantial than thin inflatables. Each material involved tradeoffs. Latex could produce flexible skins but aged, tore, and posed allergy concerns for some users. Foam could add volume but complicated cleaning and repair. Rigid internal forms improved shape while reducing flexibility.

Silicone elastomers offered a different combination of molding detail, elasticity, pigmenting, and durable surface. Their broader industrial and medical histories predated sex-doll adoption. In doll manufacturing, silicone did not act alone: sculpted masters, negative molds, internal skeletons, casting technique, seam control, hair insertion, eyes, pigments, and modular replacement systems all mattered. [S030] [S004] [S005]

The shift from an inflatable envelope to a heavy articulated body changed the value chain. It required more material, skilled sculpting, mold investment, metalwork, curing, quality control, crating, freight, maintenance guidance, and repair services. Products became customizable and capital intensive. Buyers faced new information problems involving weight, joint limits, storage, skin damage, compatibility of cleaners, replacement parts, and the truth of material claims.

“Silicone” should not be treated as a universal safety certificate. Formulations, additives, cure systems, manufacturing quality, and intended contact differ. “Medical grade” has specific regulatory meanings in some contexts and is often used loosely in marketing. No single global label makes every silicone doll equivalent. [S067] [S069] [S071]

RealDoll and the premium silicone platform #

Matt McMullen and Abyss Creations developed RealDoll in the 1990s, with company history commonly dating commercial launch to 1996. The platform combined sculpted silicone bodies, articulated internal skeletons, customizable heads and features, and a premium made-to-order business model. It became a major media reference point for the “lifelike sex doll” and influenced competitors, documentary photography, owner communities, and later robotic development. [S004] [S005] [S056] [S057]

Its historical importance does not depend on declaring it first. Rubber women, inflatables, textile figures, art dolls, and Japanese specialist products preceded it. RealDoll’s significance lies in the convergence of materials, customization, brand identity, high price, direct sales, and sustained public visibility.

The company’s own site and archived pages are primary commercial sources. They document how the brand described itself, but independent scholarship is needed for market-wide claims. Launch dates, product specifications, ownership, and availability should be checked against archived snapshots and dated reporting rather than a current marketing page alone.

RealDoll also created an expandable platform. Heads and bodies could be configured, repaired, photographed, and later connected conceptually or physically to conversational and robotic systems. That modularity made the brand central to the transition from inert doll to embodied interactive product.

Design is not a march toward realism #

Histories often describe progress as movement from crude inflatable to perfectly realistic woman. That assumes one user, one body ideal, and one purpose. Doll design has instead moved along multiple axes:

• realism versus stylization;
• full body versus torso or modular form;
• lightness versus structural durability;
• poseability versus simplicity;
• sexual anatomy versus face and character;
• generic body versus customization;
• private storage versus display;
• physical use versus photography or companionship;
• human resemblance versus fantasy aesthetics.

Realism is culturally produced. Skin color, facial proportion, hair, cosmetics, age coding, body size, and expression draw from beauty industries, pornography, fashion dolls, animation, medical modeling, and sculpture. A product can look “realistic” to one audience and stylized to another.

Design also encodes assumptions about the imagined owner. Early advertising often centered heterosexual men and female-form dolls. Modern catalogs show a wider range of body forms and customization. Those current catalogs require product-level review and do not establish user demographics, inclusive design, or the disappearance of stigma.

Owners, use, and the limits of stereotype #

The dominant media stereotype depicts the sex-doll owner as a lonely heterosexual man who has failed at relationships. Limited studies and interviews document several uses—including sexual use, companionship, photography, collecting, and aesthetic interest—but their small or self-selected samples cannot establish population prevalence or a universal owner profile. [S004] [S026] [S103] [S104]

A doll does not dictate one relationship. Within the limited published samples, some participants describe naming, dressing, photographing, or treating a doll as a companion, while others emphasize private and instrumental use. These findings describe selected participants, not all owners.

Historical interpretation must also distinguish testimony from metaphor. An owner may describe a doll as a partner while knowing it is an object. A journalist may interpret care as delusion when the same symbolic attachment to cars, instruments, heirlooms, or fictional characters would be described less pathologically. Conversely, romantic language should not be taken as evidence that an object possesses agency.

The best generalization is modest: sex dolls can participate in sexual, emotional, aesthetic, and practical routines, and the relative importance of those functions differs by person and period.

Companionship, grief, and loneliness #

Artificial-partner narratives repeatedly attach dolls to absence: sailors without women, Kokoschka after Alma Mahler, bereaved or isolated owners, and societies said to be losing intimacy. This makes loneliness both a real possible motive and a cultural explanation imposed from outside.

Evidence should be separated into three levels. DOCUMENTED individual account means a person describes companionship or grief. PROBABLE broader pattern requires multiple studies or sources. UNSUPPORTED social diagnosis claims that a whole nation or generation has turned to dolls because human relationships are collapsing.

Doll companionship is not identical to conversational AI companionship. A static body offers presence, touch, clothing, photography, spatial occupation, and routine without dialogue. A chatbot offers language and responsiveness without physical embodiment. Combined systems seek to merge those affordances, but their data dependencies and commercial control introduce new vulnerabilities. [S011] [S100] [S101]

Sex dolls, feminism, and competing critiques #

There has never been one feminist position on sex dolls. Critical arguments hold that female-form dolls objectify women, commodify sexual access, reproduce narrow body ideals, normalize asymmetrical relationships, or extend pornographic scripts into material form. Richardson’s work on sex robots represents one influential version of this critique. [S007] [S008]

Other feminist, sex-positive, and design perspectives resist treating an object’s meaning as fixed. They ask who designed it, who controls it, which bodies are represented, whether users can adapt it, and whether criticism reproduces shame around masturbation, nonpartnered sex, or disability. Feminist sex-toy retail demonstrated that product education and design could be reorganized around different users even when dolls remained marginal to many stores. [S032]

A historically careful section must distinguish empirical claim from moral argument. “This design depicts a passive feminized body” is observable. “Its use necessarily causes violence” is a causal claim requiring evidence. “The product expresses a commercial fantasy of compliance” is an interpretation. Each may be debated, but they are not the same kind of proposition.

The history also needs labor. Dolls are produced by sculptors, mold makers, factory workers, photographers, customer-service staff, repair technicians, software developers, and logistics workers. A focus only on symbolic “woman as object” can obscure the real people—often women—whose labor creates and markets the object.

What is a sex robot? #

A robust definition requires more than human shape and marketing. For this research, a sex robot is an embodied sexual system with powered actuation and at least some sensing, programmed response, or interactive control. The definition permits levels:

1. static doll;
2. doll with nonresponsive electronics;
3. remote-controlled mechanism;
4. sensor-responsive robot;
5. conversational embodied platform;
6. autonomous adaptive companion.

These levels prevent a product with a speaker or heater from being described as an autonomous robot. They also separate a robotic head attached to a static body from a full-body machine. [S004] [S005] [S009]

There is no universally accepted threshold. Engineers may define a robot through sensing, computation, and action; journalists may use the word for any interactive doll; companies may prefer it because it attracts attention. A historical timeline must state which definition it uses.

Roxxxy and the problem of the publicity prototype #

TrueCompanion introduced Roxxxy at an adult-industry trade show in 2010 and promoted it as an interactive sex robot with speech, sensors, and selectable personalities. Contemporary reporting made it a global symbol of the approaching sex-robot future. [S060] [S061]

The event proves that a company publicly presented and marketed a prototype claim. It does not by itself prove autonomous function, independent technical validation, delivery to customers, production volume, or long-term commercialization. Archived company pages are valuable for feature claims but cannot verify themselves.

Roxxxy belongs in the chronology as an early widely publicized sex-robot prototype and marketing milestone, not an undisputed first successful commercial robot. This distinction—announcement versus demonstrated system versus delivered product—is essential across technology history.

Harmony, Realbotix, and embodied conversational systems #

In 2017, Realbotix and related RealDoll projects promoted Harmony as a combination of conversational software, customizable personality, and robotic head or body components. The program drew on RealDoll’s existing body platform and the broader consumer availability of speech recognition, mobile apps, cloud services, and machine-learning branding. [S004] [S005] [S058]

Its historical importance lies in convergence. Rather than building a robot body from scratch, the system added software and mechatronics to an established doll product ecosystem. That platform logic resembles other connected consumer products: hardware becomes one layer of an ongoing service.

Current capability, pricing, delivery, ownership, and data practices are LIVE-CHECK REQUIRED AT PUBLICATION. A video demonstration can show movement or dialogue but not reliability, autonomy, security, or user experience across time. “AI” should be unpacked into concrete functions: scripted response, language model, speech recognition, memory, personalization, sensor input, motor control, and network dependence.

Are AI companions sex dolls? #

A disembodied chatbot can support romantic or sexual interaction, but it lacks the material body central to the working definition of a doll. It is better classified as an AI companion or relational agent. It enters sex-doll history when integrated into a physical doll, when users map a software persona onto a body, or when companies market cross-platform embodiment.

The boundary matters because the risks differ. A static doll has material, mechanical, storage, and hygiene issues. A cloud companion has privacy, subscription, moderation, memory, and platform-dependency issues. An embodied connected system has both.

The language of “digital sex doll” may be rhetorically useful but should not erase those distinctions. Digisexuality scholarship offers a broader category for sexual technologies that mediate or constitute sexual experience, while sex-doll history remains anchored in artificial bodies. [S011]

Privacy, cybersecurity, and the connected body #

An inert doll does not collect data. A connected doll or robotic head may process voice, images, account credentials, device identifiers, preferences, location, payment data, or intimate interaction logs. The risk depends on architecture: local versus cloud processing, encryption, authentication, update policy, third-party analytics, retention, and whether remote control is possible.

Current connected-device assessment should use a threat model rather than a sensational headline. Questions include:

• What data are collected, and are they necessary?
• Are microphones or cameras continuously active?
• Can another person control movement remotely?
• How are accounts authenticated?
• Does the product receive signed security updates?
• Can the company change functions through a subscription?
• What happens if servers close?
• Can a user delete data and export settings?

General privacy and IoT frameworks from the FTC, NIST, GDPR, and state law can guide analysis, but they do not certify any product. [S080] [S081] [S082] [S083] [S084]

All current product-specific claims are LIVE-CHECK REQUIRED because apps, ownership, policies, and software change rapidly.

Manufacturing geography and the value chain #

The modern sex-doll value chain may include concept art, body scanning, sculpting, CAD, mold fabrication, polymer compounding, casting or molding, skeleton manufacture, hair and eye components, assembly, photography, web retail, payment processing, freight, customs, customer service, repair, and replacement parts.

Different eras organized those stages differently:

• artisanal figures concentrated design and fabrication in one workshop;
• inflatable rubber goods relied on rubber processors, sheet makers, seam work, and discreet sellers;
• postwar vinyl separated mass production from branded novelty distribution;
• premium silicone firms integrated sculpture, molds, skeletons, customization, and direct sales;
• e-commerce enabled factories to sell through multiple labels;
• robotic systems added software vendors, cloud infrastructure, sensors, security updates, and subscription services.

Repair, maintenance, and product life #

Dolls are not only purchased; they are maintained. Inflatable products require patching and valve repair. Latex and vinyl age. Silicone and TPE can tear around joints or high-stress areas. Skeletons loosen or break. Hair, cosmetics, pigmentation, and surface finish change. Weight makes storage and movement part of design.

Repair creates secondary markets and knowledge communities. Owners exchange instructions, compare formulations, document defects, and identify factories behind rebranded products. This shifts authority from manufacturer copy toward user-generated evidence, though online advice varies in quality.

Modular heads, inserts, hands, and replacement parts change both economics and identity. A doll may persist through component replacement rather than remain one fixed object. Robotic platforms intensify this issue because software support can end while the body remains.

Photography, art, and the gaze #

Sex dolls appear in photography, gallery art, documentary film, horror, comedy, and science-fiction. Elena Dorfman’s Still Lovers portrayed owners and dolls in domestic settings, contributing to a visual history distinct from product advertising. Bellmer’s dolls belong to modern art; Kokoschka’s commission belongs to modernism and biography; contemporary artists use dolls to examine gender, aging, labor, simulation, and intimacy. [S019] [S022] [S026]

Images are evidence of representation, not direct evidence of prevalence. A photographer selects subjects and poses. A film may use a doll as comic shorthand for social failure. News imagery may emphasize uncanniness. Product photography constructs an ideal body and hides seams, weight, maintenance, or scale.

A publication can often explain the history without explicit images. Where images are useful, rights must be checked at the level of the photograph, not merely the age of the depicted object. The Met’s open-access Gérôme painting can illustrate Pygmalion reception, while Kokoschka, Bellmer, modern brand, and owner photographs may remain copyrighted or require permission. [S044] [S089] [S095] [S096]

Film and mass media #

Mass media helped turn the inflatable doll into a joke and the lifelike doll into a symbol of alienation, male fantasy, technological anxiety, or artificial companionship. Comedy often depends on public recognition of the object while denying owners complexity. Horror and science fiction animate the passive body, reversing control. Documentaries promise access to hidden lives but may select the most visually dramatic cases.

A historical film section should focus on shifts in representation rather than compile appearances. Useful questions are:

• Is the doll a prop, character, substitute, threat, or companion?
• Is ownership coded as comic, deviant, disabled, bereaved, violent, or technologically advanced?
• Does the work give the artificial partner agency?
• Does it distinguish an inert doll from a robot?
• How does censorship shape what can be shown?

Media visibility can normalize a category while reinforcing stigma. Those processes can occur simultaneously.

Myths and evidence: the principal corrections #

“Sailors invented sex dolls.” #

UNSUPPORTED as a documented origin. The story is historically influential but lacks the primary evidence needed for an invention claim. [S001]

“The first sex doll was Pygmalion’s statue.” #

CATEGORY ERROR. Pygmalion is literary myth, not an archaeological or commercial record. [S033]

“Goodyear invented the rubber sex doll.” #

UNSUPPORTED. Vulcanization was enabling technology; purpose-built commercial rubber women appear later. [S049] [S001]

“The Nazis created the first sex doll.” #

UNSUPPORTED / likely hoax. No credible wartime project archive supports the Borghild story. [S001] [S003]

“Bild Lilli was a Nazi sex doll.” #

FALSE. It was a postwar West German novelty/fashion doll and an influence on Barbie. [S045] [S046]

“Kokoschka invented the modern sex doll.” #

MISLEADING. His 1919 commission is documented but bespoke and artistic; commercial rubber dolls were already known. [S022]

“Bellmer’s dolls were sex toys.” #

UNSUPPORTED as functional classification. They are eroticized Surrealist artworks. [S019] [S020]

“Inflatable dolls began in the 1960s.” #

FALSE as an origin claim. Postwar plastics expanded a category documented in rubber by the late nineteenth century. [S001]

“RealDoll was the first lifelike doll.” #

CONTESTED. It was a major premium silicone milestone, not an undisputed first. [S004] [S056]

“Roxxxy was the first autonomous sex robot.” #

CONTESTED. A public launch is documented; autonomy, delivery, and commercial success require separate proof. [S060] [S061]

“Sex robots are already indistinguishable from people.” #

UNSUPPORTED marketing/futurist claim. Current systems should be evaluated function by function. [S004] [S005]

“Dolls either cause violence or prevent it.” #

INSUFFICIENT EVIDENCE. Broad causal claims exceed the available research. [S004] [S010]

“All owners are lonely heterosexual men.” #

UNSUPPORTED as a universal profile. Limited, self-selected samples document more than one motive and pattern of use; they do not establish the demographics or prevalence of all owners. [S004] [S103] [S104]

“Silicone automatically means medical grade.” #

FALSE. Material family, formulation, testing, and regulatory use must be specified. [S067] [S069]

“AI companions are the same thing as sex dolls.” #

CATEGORY DISPUTE. Embodiment, materiality, and data systems create distinct histories. [S011]

When did sex dolls become mainstream? #

Mainstreaming occurred in stages rather than one date.

1. Late-nineteenth-century specialist commerce: purpose-built rubber women circulated as discreet commodities.
2. Postwar novelty distribution: inexpensive inflatables reached mail order, adult shops, magazines, and gag-gift culture.
3. Specialist lifelike manufacturing: Japanese love-doll firms and premium Western makers developed sustained brands and owner communities.
4. Internet direct sales: anonymity, product images, forums, reviews, and international shipping expanded access and information.
5. Documentary and entertainment visibility: dolls became recognizable symbols in global media.
6. Robotics and AI discourse: the category entered technology conferences, ethics debates, and mainstream reporting.

Mainstream visibility does not mean universal acceptance. Stigma, legal restriction, payment barriers, and platform moderation remain uneven.

What sex-doll history reveals #

Sex-doll history reveals how technologies become categories through language, law, retail, and narrative. Rubber did not merely make a softer body; it created an inflatable commodity that could be hidden, shipped, and repaired. Plastic did not merely lower cost; it joined adult commerce to postwar mass manufacturing. Silicone did not merely increase realism; it created a premium customizable platform and repair economy. Robotics did not merely animate a doll; it added software claims, data collection, subscriptions, and cybersecurity.

The history also shows how artificial bodies carry social assumptions. Female-form dolls have been used to debate objectification, loneliness, prostitution, disability, gender, masculinity, and technological control. Those debates sometimes say more about the people imagining users than about documented use.

The strongest historical conclusion is therefore not a name or date. It is an evidentiary rule:

An artificial human form becomes part of sex-doll history through documented purpose, use, or commercial context—not through shape, animation, or modern analogy alone.

Patents, prototypes, and the false precision of “first” #

Patent databases can make invention history appear cleaner than it is. A filing has an inventor, date, drawings, and claims, while private practice and illicit commerce often do not. That documentary neatness tempts writers to identify the earliest matching title and call it the invention of the sex doll.

Three controls are necessary. First, patent language is strategic. An applicant may describe a figure as a mannequin, therapeutic aid, inflatable doll, novelty, anatomical model, or robot to fit a legal category or avoid explicitness. Second, a patent can claim a design never manufactured. Third, a commercial product may rely on ordinary materials and manufacturing methods without a sex-doll-specific patent.

The patent chronology should therefore record separate events:

• application date;
• priority date;
• publication or grant;
• claimed purpose;
• evidence of a prototype;
• evidence of manufacture;
• evidence of sale;
• evidence of delivery and sustained support.

Goodyear’s 1844 rubber patent belongs to enabling technology, not to the product category. The 1961 doll-construction patent associated with Barbie belongs to fashion-doll engineering, not sexual use. Later patent searches for inflatable, articulated, robotic, and interactive figures reveal a dense neighborhood of adjacent inventions whose function must be read claim by claim. [S048] [S049] [S064] [S065] [S066] [S108]

For robots, “first” becomes even more unstable. A patent may predate a working prototype; a prototype may never reach buyers; a delivered device may not perform as advertised; software may be withdrawn. The publication’s patent register therefore refuses to treat filing as commercialization.

Frequently asked questions #

Who invented the sex doll? #

No single person did. Artificial-partner myths, private commissions, commercial rubber women, postwar inflatables, silicone bodies, and robotic systems mark different thresholds. The earliest securely documented commercial class in this research is the late-nineteenth-century French femme en caoutchouc. [S001] [S037]

When were sex dolls invented? #

Stories about artificial partners are ancient, but stories are not product records. Purpose-built commercial inflatable rubber women are documented in late-nineteenth-century France; inexpensive plastic inflatables spread after the Second World War. [S001] [S002] [S037]

What is the oldest known sex doll? #

There is no uncontested oldest surviving object. Poorly preserved private figures may have existed, but the evidence does not justify assigning one ancient artifact, person, or culture the title.

Did sailors invent sex dolls? #

The sailor-made dames de voyage story is influential but lacks the contemporaneous logs, diaries, inventories, court files, or surviving objects needed to establish an origin. [S001]

Was the Borghild Project real? #

No credible wartime archive, procurement record, patent, manufacturer record, photograph, or surviving object substantiates the alleged Nazi project. It should be treated as an unsupported modern story. [S001] [S003]

When did blow-up dolls appear? #

Inflatable rubber women were sold by the late nineteenth century. Postwar vinyl and other plastics later made foldable dolls cheaper to manufacture, ship, and sell at much larger scale. [S001] [S002] [S030]

Was RealDoll the first realistic sex doll? #

Not in any uncontested sense. RealDoll was a major mid-1990s milestone in premium articulated silicone bodies, but “realistic” is subjective and earlier makers also pursued lifelike form. [S004] [S056]

What makes a sex robot different from a doll? #

This research requires powered actuation plus sensing, programmed response, or interactive control. A static body with a heater or prerecorded audio does not automatically become an autonomous robot. [S004] [S009]

Are AI companions sex dolls? #

Usually not. A disembodied chatbot is an adjacent form of artificial companionship; it enters this history when software is integrated with, or organized around, an embodied doll or robot platform. [S011]

Conclusion #

No single inventor stands at the beginning of sex-doll history. Ancient myths and later fiction imagined artificial partners; automata and mannequins supplied technologies and forms; late-nineteenth-century French sellers commercialized inflatable rubber women; twentieth-century artists built singular eroticized bodies; postwar plastics created a cheap novelty market; Japanese firms and premium silicone makers developed durable specialist industries; and robotics companies later added motors, sensors, dialogue, and AI branding.

The most famous origin stories are often the least secure. Sailor-made dames de voyage remain unproven, the Borghild Project lacks credible wartime evidence, Bild Lilli was not a Nazi sex doll, and a public robot demonstration is not the same as an autonomous product delivered at scale.

A serious history keeps uncertainty visible. It distinguishes myth from artifact, private commission from mass commodity, art doll from sex doll, patent from commercialization, static body from robot, chatbot from embodied companion, and marketing from tested capability. That structure produces a more accurate answer to the question “who invented the sex doll?”: many makers, industries, and users repeatedly created artificial bodies, but the modern commercial category emerged through industrial rubber, plastics, specialist retail, silicone, global manufacturing, and software—not one isolated invention.