内容来源：https://www.wired.com/story/goose-a-new-gay-dating-app-appears-to-be-a-psyop/

内容总结：

近日，一款名为Goose的男同交友应用因涉嫌使用AI生成的虚假网红账号进行推广而引发争议。该应用自称是“Grindr的替代品”，旨在帮助男同性恋者建立长期关系，但其推广手段被指“具有欺骗性”。

据调查，多个Instagram账号如@miles.sumrall和@danielmmulugeta，在“亲密好友”故事中发布相同文案，邀请用户加入Goose“私密社群”。然而，这些账号均于2026年5月创建，发帖极少，且关注者数量远高于粉丝数。AI检测软件显示，其头像有超过90%的概率为AI生成，谷歌Gemini的SynthID检测也证实其“大部分或全部”使用了谷歌AI工具。

这些账号不仅通过私信向用户发送邀请码，还互相点赞评论以营造真实感。营销从业者Ryan Cheam表示，他收到邀请后因账号可疑而产生怀疑。另一用户Dalton Bauer则在一周内接连收到三条一模一样的话术私信，直言“这太可疑了，需要被曝光”。

Goose由模特兼网红Derek Chadwick及前BeReal运营人员David Aliagas共同创立。Aliagas曾在Instagram故事中公开招募“大使”，要求对方每天管理3个Instagram账号、工作4小时，为期2至3个月，月薪1800至2100美元，并明确表示“正在购买假账号（finstas），每个100美元”。该应用上线后一度冲上苹果应用商店免费生活类下载榜第四位。

广告与电商法律专家Rob Freund指出，若品牌创建虚假账号伪装成真实用户以推广产品，即使应用免费，也明显违反美国联邦贸易委员会（FTC）关于禁止虚假广告的规定。Meta虽要求用户标记AI生成内容，但对私信和“亲密好友”故事这类私密渠道的监管难度较大。目前Goose方面未回应置评请求，FTC也拒绝就具体案例发表评论。

中文翻译：

Instagram账号@miles.sumrall的“挚友快拍”中，一名卷曲深色头发、留着精心修剪胡须的亲切男子正漂浮在水面上，笑容满面。“你收到这条消息，是因为你正是我们打造这款产品的目标用户。”配文写道，随附一个“仅限会员社区”的邀请码。

该链接导向名为Goose的男同交友与社交应用登录页面，其口号是“专为男孩打造”。据官网介绍，该应用让用户能“通过现有生活结识志同道合的人”。

问题在于，@miles.sumrall这个账号似乎并不真实。同样不真实的还有@danielmmulugeta——那位分享上述配文、用词完全一致的深色头发帅气博主，他的内容同样出现在“挚友快拍”中。这两个账号均创建于2026年5月，帖文不足10条，且关注数远超粉丝数。根据AI图像检测软件分析，两人Instagram头像被判定为AI生成的可信度超过90%。谷歌双子星（Gemini）的SynthID检测功能也发现，迈尔斯和丹尼尔的头像照片“全部或大部分”由谷歌AI生成。

由模特兼博主德里克·查德威克与BeReal前增长及社区经理大卫·阿利亚加斯共同创立的Goose，定位为Grindr的替代品，专为寻求稳定关系的男同群体设计。该应用发布时，许多人对它能否摆脱“约炮工具”定位嗤之以鼻。“Goose本质上就是宝可梦约炮版。”一位X用户调侃道。

尽管如此，用户兴趣显然高涨——该应用上周四上线后，迅速攀升至苹果应用商店免费生活类应用下载榜第四位，目前位列全球生活类应用下载第33名。而@miles.sumrall等博主的推广内容，很可能在推动大量下载中发挥了作用。

迈尔斯和丹尼尔的截图出现在X用户@pspthe2nd的帖文中，该帖子指控该应用“使用AI模型推广虚假关注度#goose”。但这两个账号似乎只是庞大网络中的一环——众多面容英俊、疑似AI生成的男博主正通过私信或添加“挚友快拍”的方式推广该应用。

市场营销与公关客户经理瑞安·切姆表示，约一周前他首次注意到名为@alistaircrombbie的奇怪新Instagram账号。简介写着他在知名画廊从事公关工作，切姆告诉《连线》杂志：“我以为他只是个普通基友。”但当阿利斯泰尔发来私信，邀请他加入Goose的“精选男性社群”并附上邀请码时，他起了疑心。SynthID检测发现，阿利斯泰尔的头像“全部或大部分”由谷歌AI生成。

除迈尔斯、阿利斯泰尔和丹尼尔外，《连线》杂志还识别出二十余个类似账号，均创建于2026年5月或6月，帖文寥寥——这是非真实账号的典型特征。许多账号还频繁互评照片，使用相同的心形和火焰表情符号。

这些账号通常先关注潜在用户，将其加入“挚友快拍”，有时会直接发送私信鼓励注册。从事市场营销的道尔顿·鲍尔就收到了@lucalepkowski发来的私信：“嘿！这可能有点突然，但感觉你会感兴趣：)”随后用与切姆收到的完全相同的措辞邀请他加入Goose社群。

鲍尔对此并不领情：这是他本周第三次收到相同措辞的私信，均来自全新账号。“这是我第一次在Instagram看到这种规模的操作，”他告诉《连线》杂志，“这种行为十分可疑且具有欺骗性，必须有人曝光。”

与Goose邀请相关的其他账号类似，@lucalepkowski创建于2026年5月。AI图像检测软件分析显示，其头像——一名穿卡其短裤的大学生模样男子在海滩上啜饮大水瓶——有80%可能性为AI生成。谷歌双子星进一步分析确定，卢卡头像照片“至少有部分被谷歌AI编辑或生成”。

查德威克及Goose其他成员未回应多次置评请求。但联合创始人阿利亚加斯似乎在其Instagram快拍中发布了多个“大使”岗位招聘，协助管理匿名社交媒体账号。

六周前阿利亚加斯发布的快拍配文写道：“新应用需要帮手，你们知道我总会优先给老粉机会：)”他列出“大使角色”详情：每天管理三个Instagram账号四小时，持续两个月以上，月薪1800至2100美元。“熟悉同性恋文化是加分项。是时候把你的创伤变现了：)”阿利亚加斯写道，“顺便说一句，我还在以100美元一个的价格收购小号：)。”

三周前，阿利亚加斯再次为“新应用”招募“大使”，这次要求三个月任期。“我们要大干一场：-)”他写道。两条快拍仍保留在Instagram“大使”合集标签下。

品牌与开发者秘密使用AI生成的博主推广产品并不罕见。近期《卫报》调查引用一位前经纪人（如今运营此类账号）的估算：“大品牌推广内容中40%到60%是AI生成的”，且多数未公开标注。

广告与电商律师罗布·弗罗因德表示，这与“内容合规”并非同一概念。美国联邦贸易委员会（FTC）禁止欺骗性广告，例如品牌用AI冒充真实人物。纽约州近期也立法要求广告商披露AI生成内容，违者面临初始1000美元罚款。

“若创建虚假账号推广产品，或批量制造看似真实用户的假账号以吸引关注度或销售额，这种行为明显违反FTC指南。”弗罗因德强调，即使应用免费也不例外。FTC发言人拒绝回应《连线》提问，称“无法评论特定公司行为”。

Instagram母公司Meta拒绝就本文置评。根据其内容准则，用户须标注AI生成内容，违规帖文将被删除。但私信与“挚友快拍”的私密性可能增加平台监管难度。正如《连线》此前报道，错误标注的AI生成内容常能蒙混过关，让用户相信@miles.sumrall和@lucalepkowski这类假账号是真人。

不过，显然有人并不上当。“一方面，我很荣幸成为他们的目标用户，”切姆告诉《连线》，“但用这种方式引诱基友注册，实在令人不齿。”

英文来源：

The Instagram Close Friends Story for @miles.sumrall shows an affable-looking guy with curly dark hair and an expertly groomed mustache beaming as he floats on the water. “You’re receiving this because you’re exactly the type of person we’re building this for,” the caption reads, accompanied by a code for an invite to a “members only community.”
The link leads to a login for Goose, a dating and friendship app for gay men with the slogan “for the boys,” which allows users to “meet guys through the life you already have,” according to its website.
The problem is that @miles.sumrall does not appear to be real. Neither does @danielmmulugeta, the cute dark-haired influencer who shared the above caption, with the exact same verbiage, on his Close Friends’ Stories. Both accounts were created in May 2026 and have fewer than 10 posts as well as a high following-to-follower ratio. And both of their Instagram avatars were determined with greater than 90 percent confidence to be AI-generated, according to the AI Image Detector software. A SynthID check on Google Gemini, which can help identify AI-generated images, also found that “most or all of” Miles’ and Daniel’s profile photos were created using Google AI.
Created by the model-influencer Derek Chadwick, as well as former BeReal growth and community manager David Aliagas, Goose positions itself as a Grindr alternative for gay men who want to build lasting relationships. At the time that it was announced, many scoffed at the idea that the app would be used for anything other than finding casual hookups. “Goose is basically Pokémon Ho,” one X user joked.
Still, user interest was apparently high enough that when the app launched last Thursday, it rose to #4 in the App Store’s free lifestyle downloads category, and is now ranked 33rd in lifestyle app downloads globally. And promotional content by creators like @miles.sumrall likely played a role in driving so many to download the app.
Miles and Daniel appeared in screengrabs shared on X by user @pspthe2nd, whose post alleged that the app “use[s] AI models to promote fake interest #goose.” But both of the accounts appear to be part of a much larger network of comely, seemingly AI-generated male influencers promoting the app, either by reaching out to gay men via DM or adding them to their Close Friends Stories.
Ryan Cheam, an account executive in marketing and public relations, says he first noticed a strange new Instagram account belonging to someone named @alistaircrombbie about a week ago. His bio says he works in PR at a well-known art gallery, so, Cheam tells WIRED, “I thought he was just a normal gay guy.” He became suspicious, however, after Alistair DMed him inviting him to join a “curated network of guys” at Goose, sending him an invite code. A SynthID check found that “most or all” of Alistair’s profile photo was generated using Google AI.
In addition to Miles, Alistair, and Daniel, WIRED was able to identify more than two dozen similar accounts, all of which were created in May or June 2026 and featured just a few posts—a typical indication of inauthentic accounts. Many of the accounts also frequently comment on each other’s photos, including the same heart and fire emojis.
Often, the accounts followed potential members and added them to their Close Friends Stories, but sometimes they directly DMed them to encourage them to sign up, as was the case with Dalton Bauer, who works in marketing and received a DM from a user named @lucalepkowski. “Hey! Okay this might feel random but felt you’d be interested :),” the message begins before inviting Bauer to the Goose community, using language identical to that of the one Cheam received from Alistair.
Bauer was less than flattered by the invite: it was the third such DM he’d received this week, using the exact same language, from an account that was brand-new. “This is the first time I’ve seen this on Instagram, and at this scale,” he tells WIRED. “I think someone needs to shed light on this as it’s shady and deceiving.”
Like the other accounts linked to the Goose invites, the @lucalepkowski account was created in May 2026. An analysis using the AI Image Detector software found that the profile image—a photo of a college-aged man in khaki shorts on a beach, sipping from a large water bottle—was 80 percent likely to have been artificially generated. Further analysis using Google Gemini also determined that at least part of Luca’s profile photo “was edited or generated with Google AI.”
Chadwick did not respond to multiple requests for comment, nor did anyone else from Goose. But on his Instagram Stories, cofounder Aliagas appears to have advertised multiple job listings for “ambassadors” to help manage various unspecified social media accounts.
The caption of an Instagram story Aliagas posted six weeks ago reads: “Need some help w my new app and you know I always give priority access to these opportunities to my OGs here :)” He listed the details for the “ambassador role,” requiring that they help manage three Instagram accounts four hours a day for two-plus months, with rates ranging from $1,800 to $2,100 a month. “Familiarity with gay culture is a big +. Time to monetize ur traumas :),” Aliagas wrote before signing off with, “And btw, still buying finstas [fake Instagram accounts] for $100 :).”
Aliagas posted another call for applicants for an “ambassador role” for his “new app” three weeks ago, this time asking for a three-month commitment. “We are going big :-),” he wrote. Both Stories are still included in his Instagram Story highlights, under the title “AMBASSADORS.”
It’s not unusual for brands and developers to surreptitiously use AI-generated influencers to promote their products. A recent Guardian investigation quoted a former celebrity manager who now creates such accounts, who estimated that “40 percent to 60 percent of the content out there from some of the big brands” is artificially generated, most of which is not publicly identified as such.
That’s not the same thing, however, as saying such content is aboveboard, says Rob Freund, an advertising and ecommerce attorney. The US Federal Trade Commission has guidelines prohibiting deceptive advertising, such as brands using AI to impersonate real individuals. New York also recently enacted a law requiring that advertisers disclose if their content is AI-generated, and companies face an initial $1,000 fine if they fail to do so.
“If you are creating fake accounts for people who promote a product and explicitly creating a bunch of fake accounts that look like they are users of a product or a service to drive attention or sales to that product or service, that activity is very obviously unlawful under FTC guidelines,” says Freund, stressing this is still the case regardless of whether the app is free. An FTC spokesperson declined to respond to questions from WIRED, saying it “cannot comment on a specific company’s practices.”
According to its content guidelines, Meta, which owns Instagram and declined to comment for this story, requires users to label their content as artificially generated and will take down posts that are not correctly identified as such. But the private nature of this specific marketing campaign in DMs and Close Friends Stories may make it more difficult for the platform to regulate. And as WIRED previously reported, incorrectly labeled AI-generated content often escapes detection, allowing users to believe that accounts that are most likely fake, such as @miles.sumrall’s and @lucalepkowski’s, are the real deal.
Some people, however, clearly aren’t falling for it. “On one hand I’m flattered that I’m their target audience,” Cheam tells WIRED. “But the need to essentially bait gay guys into signing up feels really sketchy.”