18.03.2026 12:15Author: Viacheslav Vasipenok

How Higgsfield AI Became 'Shitsfield AI': A Cautionary Tale of Overzealous Growth Hacking

News image

Once hailed as the fastest-growing AI startup in the world, Higgsfield AI has plummeted from darling of the tech scene to a punchline for hype gone wrong. Founded in 2023, the San Francisco-based company exploded onto the generative video market with promises of effortless, high-quality content creation. By early 2026, it boasted 15 million creators producing 4.5 million clips daily and a staggering $300 million annual recurring revenue (ARR).

Yet, aggressive marketing tactics, including distributing racist videos and botching influencer payments, have tarnished its reputation. Derisively nicknamed "Shitsfield AI" by critics, the startup now faces accusations of empty hype and unethical practices. This case study dissects how a promising venture overreached, alienating influencers and users alike, and what it means for the AI industry's growth-at-all-costs mentality.


The Market Landscape and Higgsfield's Meteoric Rise

In 2023, the generative video space was nascent and fragmented. Heavyweights like Google and OpenAI were still testing closed betas, while early players like Pika offered primitive tools. Runway stood out for quality but required steep learning curves. Enter Higgsfield AI, co-founded by Alex Mashrabov — a Russian émigré and former Snap engineer — and Yerzat Dulat, a Kazakh AI specialist.

Their web app simplified video generation: users inputted text prompts, product links, or images, and the system leveraged 12 top-tier models (including being OpenAI's largest Sora 2 customer) with 400 presets for effects and motions.

The killer feature? One-button UX tailored for marketers, SMM pros, and casual creators needing "good enough" videos fast for social media. Higgsfield didn't openly disclose relying on open-source models, and no proprietary tech was publicly verified, but that didn't matter—adoption soared. By January 2026, it raised $80 million from Accel, GFT Ventures, and Menlo Ventures, signaling investor confidence in its pivot from a mobile app to serving ad agencies and influencers.


Seizing the Window: A Strategy Focused on Influencers

Recognizing a narrow opportunity before Big Tech saturated the market, Higgsfield doubled down on growth hacking. Instead of targeting enterprise clients, they courted mid-tier and micro-influencers—cost-effective partners with loyal, engaged audiences. The logic was sound: AI video's viral potential didn't depend on influencer size, and controversial or eye-catching content could amplify reach algorithmically.

Through the Higgsfield Earn program, creators were incentivized to promote the tool, often with access to polished templates and pre-made videos. This approach initially worked wonders, boosting user numbers and revenue from $200 million ARR in late January to $300 million by early February 2026. With 300,000 paying subscribers, the startup seemed unstoppable.

Where It Went Wrong: Overreach and Ethical Lapses

Higgsfield's zeal crossed into problematic territory. First, influencers received overly refined templates that produced demo videos far superior to what average users could achieve. These "perfect" clips went viral, drawing in users who then faced disappointing results, leading to frustration directed at promoters rather than the company.

More damningly, the startup allegedly pushed provocative briefs. Marketing materials included racist videos — like Shrek characters using anti-Asian slurs, Moana declaring "My body is for white people," and nonconsensual deepfakes of celebrities like Sydney Sweeney and Zendaya, or even political figures like President Trump. Such content was designed to virally explode, but it backfired spectacularly, drawing ethical fire and alienating creators who faced backlash for sharing it.

Payment woes compounded the issues. In the Earn program, 90% of submissions were paid, but fraud, bots, and delays plagued the rest, with creators reporting withdrawal problems and sudden bans. Higgsfield also passed off stock footage from Envato as AI-generated, offered murky partnerships (like ads for discounts — a major faux pas), and lured users with 65% off "unlimited" plans that later throttled speeds, rendering the app unusable without extra credits. Influencers were treated as disposable: burned out quickly and replaced, without fostering long-term relationships.


The Backlash: From Hype to Hate

The strategy imploded as influencers turned vocal critics. Creator Tim Soret decried the hype as "fake and it’s bought," while Dustin Thornton slammed the "ruthless tactics." Users followed suit, churning away amid disillusionment. Media piled on, with Forbes exposing the "dark side" in February 2026, and Higgsfield's X account getting suspended for inauthentic behavior.

CEO Mashrabov acknowledged missteps: "Rapid scaling brings real challenges. We acknowledge that our internal processes and external communications haven't always kept pace with our core values, and we have made mistakes." Cofounder Mahi de Silva (noted in the article, though original founders are Mashrabov and Dulat) added, "We fully admit that we push the envelope... it's more controversial content that gets attention," calling racist videos a "mistake" unrepresentative of values. The company pledged mandatory reviews for materials.

Despite this, Higgsfield retains its valuation and metrics — for now. But positioning has eroded: from "user-friendly leader" to "toxic also-ran" in a crowded market.

Alo read:


Lessons Learned: Build Partnerships, Not Pyres

This saga underscores the perils of growth hacking without ethics. Brands must cultivate genuine, long-term influencer relationships, not exploit them as fodder. As the AI video space matures — with competitors like Runway and Big Tech entrants — trust crises can lag but ultimately erode metrics. For Higgsfield, upcoming funding rounds will test if "Shitsfield" sticks. The moral? In tech's hype machine, burning bridges burns brighter than viral flames.


0 comments
Read more