Tag: Digital Identity

Posted on by

Synthetic profile photographs and the appearance of personal identity

A face does a surprising amount of work in an online profile.

Add a photograph to a username and a thin account suddenly feels inhabited. Readers infer age, mood, style, maybe even personality. A realistic headshot can make a profile seem less like an entry in a database and more like a specific human being.

That shortcut worked reasonably well when producing a convincing portrait required access to an actual person or a skilled image editor. Generative image systems changed the cost.

A 2024 study by researchers at Indiana University examined fake social-media profiles using AI-generated faces and collected 1,420 accounts using GAN-created portraits. The researchers found those accounts participating in scams, spam, and coordinated amplification. Their paper, “Characteristics and prevalence of fake social media profiles with AI-generated faces”, estimated a lower bound of roughly 0.021% to 0.044% of active Twitter/X accounts in their sample using GAN-generated profile faces.

The percentage matters less than the mechanism. A synthetic portrait lets an operator manufacture the visual evidence of a person without borrowing a stock photograph or stealing somebody else’s selfie.

A photograph establishes appearance, not identity

Even a completely genuine photograph does not prove who controls an account. The person in the image may be the operator, a friend, a celebrity, a victim of identity theft, or somebody whose picture was copied from another site.

A generated photograph removes even that weak connection.

This is why spotting a synthetic face should be treated as evidence about the image, not a complete diagnosis of the account. Some fake accounts use real photographs. Some legitimate users use avatars, illustrations, masks, game characters, or generated artwork. An account with an AI face may be deceptive, openly fictional, experimental, or simply privacy-conscious.

The useful question is not “Does this face look real?” It is “What evidence connects this profile to the identity it claims?”

Identity has to be assembled from more than pixels

Useful evidence can include a long account history, links from an independently controlled website, consistent participation in a known community, verifiable work, cryptographic signatures, live interaction, or other people who can establish continuity over time.

None of those methods is perfect. A sophisticated fake can build history. Real people can be private. Pseudonymous users may deliberately avoid linking their legal names to public accounts.

That is precisely why the profile photograph should carry so little weight.

Synthetic faces matter because they automate one of the oldest trust cues on the web. They do not create fake people by themselves. They create something subtler: the appearance that a person has already been established, before any actual evidence of identity has been examined.

Posted on by

Shared accounts that complicate the idea of one user, one person

The internet encourages a convenient fiction: one account equals one person.

Sometimes it does. Often it does not.

A restaurant account may be handled by whoever is working that week. A newsroom account can be used by several editors. A company support profile may be staffed around the clock by different employees. Families share logins. Clubs, nonprofits, bands, open-source projects, and political organizations all maintain public identities that outlive whichever individual happens to be typing.

That makes account-level measurement a poor shortcut for counting people.

Platforms explicitly support multi-person identities

This is not an obscure edge case. Meta’s Facebook documentation says a Page can give multiple trusted people access to create content, answer messages, respond to comments, manage ads, and perform other tasks as the Page.

To the public, the Page is one identity.

Behind it may be five humans with different schedules, writing styles, locations, devices, and habits.

A behavior-based classifier watching only the public stream could interpret those changes in several ways. Posting may appear around the clock. Tone may shift abruptly. One person may write long conversational replies while another mostly posts links. A third may schedule promotional messages in advance.

The account looks inconsistent because it is not one person.

Shared identity creates strange measurement artifacts

Suppose a study counts 10,000 accounts and treats them as 10,000 users. Some accounts belong to one person. Some people own several accounts. Some accounts are operated by teams. Some are bots. Some are organizations using automation plus human staff.

The total number of actual people cannot be recovered by simply counting rows in the account table.

Shared accounts can also confuse attempts to detect automation. A corporate profile may publish on a precise schedule because software queues posts, then switch to unmistakably human conversation when an employee replies. One analyst calls it a bot. Another calls it human. Both may be looking at real parts of the same operating model.

This is one reason the phrase “fake population” needs careful handling. The account may not correspond to a single human, but that does not make it fake. A library, newspaper, game studio, or community project is a real social actor even if no individual human maps one-to-one onto its username.

Identity online is often organizational

The better question is what kind of entity the account represents and how its activity is produced.

Is it one person? A rotating staff? A human using scheduling tools? A bot controlled by a team? An organization publishing official statements? A compromised profile? Those categories are more useful than forcing everything into “human account” or “bot account.”

Dead Internet Theory is strongest when it notices that online identity is becoming synthetic, automated, and difficult to verify. But difficulty does not justify assuming every non-personal account is artificial.

Sometimes one username hides a machine.

Sometimes it hides the night shift.

Posted on by

Cyworld: digital possessions and identity across a social world’s decline and revival

For much of the 2000s, the most important room in South Korea was a virtual one. On Cyworld, a social network that began in 1999 as a project by KAIST graduate students, every member got a “minihompy”: an isometric mini-house where a cartoon “mini-me” avatar stood among wallpaper, furniture, and a background song. Friends joined each other as “ilchon” and dropped by to leave guestbook notes. How a person furnished that room was daily social information, a compact public statement of taste, mood, and status.

The minihompy launched in September 2002, and co-founder Dong-Hyung Lee later described it as a last chance to keep a near-defunct service alive. It worked. Cyworld was bought by SK Communications in 2003 and became one of the first services anywhere to profit from virtual goods. Identity had a price: users bought dotori, or acorns, for about 100 won each and spent them on wallpaper, background music, and fonts. Many items were rentals that simply expired unless repurchased. The economics were serious — roughly 80 percent of Cyworld’s Korean revenue came from decorating purchases by 2006 — and so were the numbers. Around 13 million people, more than a quarter of the country, were members by 2005, and roughly 32 million of some 50 million Koreans held accounts at the peak.

A world its users couldn’t keep

Decline was not deletion but a slow loss of access. The iPhone arrived in Korea in 2009, and Facebook, Instagram, and KakaoTalk all shipped experiences built for phones. Cyworld’s house never made the trip; observers consistently described a platform that reacted too late to mobile. In 2011 a hacking incident exposed data tied to some 26 million Cyworld accounts in what was then one of Korea’s worst breaches. Ownership passed from SK to the founder of the rival portal Freechal in 2016 without a turnaround, and the service shut down in 2019. For users who had paid for songs and sofas, the end was literal: years of photos, posts, and purchases sat behind an offline login.

The possessions were always the pitch for revival. Cyworld Z took over the service in 2021 and said the old servers still contained an enormous archive. In March 2021 the company reported that about 18 billion records from 32 million members remained intact, including roughly 17 billion photos and 150 million videos. The mobile service finally reopened in April 2022, accompanied by plans for NFTs, a metaverse and a new cryptocurrency called Dotori; contemporary coverage of the relaunch described the attempt to turn recovered nostalgia into a new digital economy. By November 2024 the business had changed hands again, to a special-purpose company called Cycomms that promised 3D mini-mes and chat — then ran out of money for data restoration.

The long, contested comeback

In August 2026 the blockchain firm Sigmachain held a Seoul press conference titled “The Return of Cyworld.” It says it has acquired Cyworld-related intellectual property, trademarks, and user data and plans an October 2026 beta centered first on friend connections and a feed. Digital Today’s report on the announcement says the company aims to restore roughly 17 billion records and launch a fuller service in the first half of 2027, while also considering NFTs and stablecoin-based features.

What a revival could genuinely recover is the archive — the rooms, songs, and photos that once stood in for their owners. What stays uncertain is whether this round holds. The announced acquisition and relaunch remain entangled in disputes over business rights, data, and control of the service, so the October beta is a plan rather than a completed restoration. The parts users could buy were always recoverable assets; even in the darkest years the archive kept its value. Ownership has been the fragile half of the arrangement all along.