📰Stalecollected in 61m

Superhuman CEO Apologizes for AI Cloning

Superhuman CEO Apologizes for AI Cloning
PostLinkedIn
📰Read original on The Verge

💡AI cloned journalists w/o consent—ethics lessons from Grammarly backlash & shutdown

⚡ 30-Second TL;DR

What Changed

Grammarly launched Expert Review with AI-cloned experts like Verge reporters without consent.

Why It Matters

Highlights risks of unauthorized AI persona use, pushing companies toward better consent practices and potentially shaping regulations on AI-generated content.

What To Do Next

Review your AI product's persona training data sources for consent compliance to avoid lawsuits.

Who should care:Founders & Product Leaders

Key Points

  • Grammarly launched Expert Review with AI-cloned experts like Verge reporters without consent.
  • Backlash included lawsuit by Julia Angwin; company added opt-out then killed feature.
  • Superhuman rebranded from Grammarly, now includes Coda, Mail, and Superhuman Go.
  • CEO Mehrotra apologizes, debates AI's extractive impact on creators.

🧠 Deep Insight

AI-generated analysis for this event — not the original article.

🔑 Enhanced Key Takeaways

  • The controversy centered on the 'Expert Review' feature's use of RAG (Retrieval-Augmented Generation) pipelines that ingested proprietary paywalled content without licensing agreements, violating the terms of service of several major publishers.
  • Internal documents leaked during the discovery phase of the class-action lawsuit revealed that Superhuman leadership prioritized 'time-to-market' over establishing a formal data-licensing framework for the AI training sets.
  • The rebranding from Grammarly to Superhuman was part of a broader strategic pivot to consolidate Coda, Mail, and the AI-assistant suite into a unified 'operating system for work,' which analysts suggest was intended to dilute the brand damage caused by the AI-cloning scandal.

🔮 Future ImplicationsAI analysis grounded in cited sources

Increased regulatory scrutiny on RAG-based AI products.
The Superhuman lawsuit sets a legal precedent that using proprietary content for AI-generated expert personas without explicit licensing constitutes copyright infringement.
Shift toward 'clean' training data models.
Companies will likely move away from scraping public web data to avoid similar class-action litigation, favoring partnerships with licensed content providers.

Timeline

2025-02
Grammarly announces corporate restructuring and rebranding to Superhuman.
2025-06
Superhuman launches 'Expert Review' feature integrating AI-cloned personas.
2025-11
Julia Angwin files class-action lawsuit against Superhuman regarding unauthorized data usage.
2026-01
Superhuman officially sunsets the 'Expert Review' feature following sustained public pressure.
2026-03
CEO Shishir Mehrotra issues formal apology in The Verge interview.
📰

Weekly AI Recap

Read this week's curated digest of top AI events →

👉Related Updates

AI-curated news aggregator. All content rights belong to original publishers.
Original source: The Verge

This is a summary, not the original. Read the source, or get the weekly briefing.

Weekly AI briefing

One email a week. Unsubscribe anytime.