Dozens of Black-Market AI Training Accounts Discovered for Sale
A new trend in the artificial intelligence sector is exposing the rise of a black market for training accounts. Major companies, including Scale AI, Surge AI, and Mercor, are relying on a global workforce for AI chatbot training. This reliance has led to a shadow economy where accounts necessary for this work are being sold illegally.
Illicit Marketplace for AI Training Accounts
An investigation by Business Insider identified over 100 Facebook groups involved in selling access to both real and fake AI training accounts. These accounts are essential for contractors who need them to work for various data-labeling firms. After the issue was reported, Meta took action, removing around 40 groups that violated their policies. Further investigations are ongoing.
Companies’ Position on Account Reselling
Leading AI training companies like Scale AI have strict rules against account reselling. Despite this, “verified” accounts continue appearing for sale across platforms such as Facebook, WhatsApp, and Telegram. Internal documents have revealed that Scale AI has struggled with fraudulent account use and VPN issues for several years.
- Scale AI received a $14 billion investment from Meta in June 2023.
- Pay for contractors can exceed $100 an hour.
- The company has barred specific countries to mitigate cheating.
Nature of AI Training Work
Contractors primarily engage in tasks like labeling images and providing chatbot responses. They must first create accounts and pass various screening tests on platforms managed by companies like Scale AI and Surge AI. When opportunities diminish, unscrupulous individuals sell accounts, often claiming them to be “verified.”
Fraudulent Practices Targeting Account Buyers
The trend of reselling accounts has raised concerns. Buyers may be scammed, risking exposure of their personal information. Conversely, sellers worry that buyers might vanish after receiving account details without compensation. Reports highlight increasing instances of scams occurring within the AI training sphere.
Scale AI’s Response to Fraud
Reports indicate that Scale AI has been combating scammers for years. Company documents show numerous instances of contractors flagged for potentially fraudulent activities. For instance, during a project for Google in 2024, a significant number of taskers were identified as suspected cheaters or spammers.
- Scale AI maintains mechanisms to detect fraudulent activities.
- They have prohibited account creation from certain regions.
- Recent documents reveal over 490 contractors were removed for violations.
Emerging Challenges Amidst the Black Market
Recent trends suggest that the urgent need for AI training and data labeling has made the market vulnerable to fraud. There is a notable arms race between AI training companies and fraudsters. Companies are urged to adopt more proactive measures to counter evolving tactics used by scammers.
As the demand for AI development continues to increase, the existence of a black market for training accounts raises critical concerns. The integrity of data labeling efforts is at risk, impacting both contractors and clients alike.
