Urgent.News

What's breaking now, across thousands of outlets.

AI

A college student thought he'd struck gold training AI. Then he says he got burned.

Nearly two dozen current and former contractors said Handshake this year underpaid them by hundreds or thousands of dollars each.

Joel Thomas, a college student, believed he had found lucrative opportunities training AI through Handshake AI, a startup that offered high-paying jobs to university students. Handshake advertised AI training work paying $50 an hour or more, which would have covered Thomas' college expenses for a year. However, he claims he is owed about $11,000 in unpaid wages for hundreds of hours of work he completed on Handshake's platform.

Handshake denied his appeal, stating there was no record of his work on one project during a specific week.

Thomas is not the only one affected. Nearly two dozen current and former contractors have reported being underpaid by hundreds or thousands of dollars each by Handshake. The total disputes from these contractors over the past six months amounted to $64,000. Handshake, which has hosted over 188,000 contractors earning over $500 million for AI training projects, stands accused of underpaying its workforce.

The underpayment cases at Handshake arose due to discrepancies in time and pay tracking, insufficient full-time staff, and pressure to deliver training data quickly. Project managers often lacked the time to address contractors' questions and concerns, leading to confusion and payment disputes. Contractors were also subject to review, and if they did not understand the instructions or the task failed the review, they did not receive payment.

Handshake claims to have introduced changes to improve pay transparency, including a new payments dashboard, clearer explanations for adjustments, and an easier dispute process. However, Thomas and other contractors remain dissatisfied, emphasizing the need for competitive pay and clear expectations.

Written by urgent.news from Business Insider's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Read the original at businessinsider.com →

More in AI

TSMC Invests $1 Billion in OpenAI, Anthropic

Taiwan’s TSMC, the world’s largest foundry (semiconductor contract manufacturing) company, will invest up to $1 billion in OpenAI and Anthropic.

  • TSMC invests up to $1 billion in OpenAI and Anthropic.
  • $500 million each for OpenAI and Anthropic approved on November 10.
  • TSMC forming investment company for $500 million in semiconductor industry.

The Tool-Design Mistakes That Make AI Agents Unreliable

The Tool-Design Mistakes That Make AI Agents Unreliable A lot of "the model is unreliable" complaints are actually tool-design problems.

  • Ambiguous tool definitions cause unreliable AI agent behavior
  • Overloaded tools with unrelated tasks confuse the model
  • Clear instructions and error handling improve AI agent reliability

AI Weekly: Fable 5.2, Muse Ember, and TPU 8i Max

Anthropic shipped Claude Fable 5.2 and cut its price by 20%. Meta launched Muse Ember, a tiny-priced model built to run the cheap steps inside agents, and Thinking Machines released Inkling 2 Small…

  • Anthropic released Claude Fable 5.2 and Mythos 5.2 on November 9
  • Fable 5.2 priced at $8 per million input tokens, $40 per million output tokens
  • Meta unveiled Muse Ember, the smallest and cheapest Muse model

More from Wednesday 11 November →