Listen To The Show
Transcript
Welcome to The Prompt by Kuro House, your daily AI update. Today, we’re diving into some big moves in AI, from legal battles to massive investments and acquisitions. Let’s get right into it.
First up, Grammarly is in hot water over its AI feature called Expert Review. According to The Verge, the company used the identities of real people, including journalists and editors, without their permission to generate AI writing suggestions. One journalist, Julia Angwin, has filed a class-action lawsuit claiming this violated privacy and publicity rights. Grammarly’s parent company, Superhuman, responded by disabling the feature and apologizing, promising to rethink how experts are represented in their AI tools. This story highlights the growing pains as AI companies navigate ethical and legal boundaries.
Continuing with Grammarly, they’ve officially announced they will stop using AI to clone experts without their consent. Superhuman CEO Shishir Mehrotra explained that the Expert Review agent was designed to help users discover influential perspectives but fell short in execution. They’re now focusing on giving experts real control over how their knowledge is represented, or whether they want to be represented at all. The company envisions a future where users have a whole team of AI agents crafted by experts, shaping the AI experience collaboratively. This is a significant pivot toward respecting expert rights while advancing AI assistance.
In other news, Nvidia is making a massive $26 billion investment to build open-weight AI models over the next five years. WIRED reports this move positions Nvidia to compete directly with AI giants like OpenAI and Anthropic by releasing models with publicly available weights and architecture. The company just launched Nemotron 3 Super, a 128-billion-parameter model that outperforms other open models on multiple benchmarks. Nvidia’s strategy includes using these models to improve not just AI software but also their hardware and datacenter infrastructure. This bold step could reshape the AI landscape by promoting openness and innovation while strengthening Nvidia’s market leadership.
On the acquisition front, Zendesk is buying Forethought, a startup specializing in AI-powered customer service automation. TechCrunch covers how Forethought, the 2018 TechCrunch Battlefield winner, has been a pioneer in agentic AI for customer interactions, supporting over a billion monthly interactions by 2025. Zendesk plans to integrate Forethought’s technology into its AI products, accelerating its roadmap by more than a year. This deal underscores the growing importance of AI agents in transforming customer experience across industries. We’ll likely see more autonomous capabilities and voice automation coming from Zendesk soon.
Finally, the legal storm around Grammarly’s AI feature is getting more attention with a class-action lawsuit filed in New York. Wired reports that the suit alleges Grammarly misappropriated the names and likenesses of hundreds of writers and experts without consent to profit commercially. Julia Angwin, the lead plaintiff, points out that the AI-generated advice was often inaccurate and misrepresented her voice. Superhuman’s CEO acknowledged the missteps and the company is now reimagining the feature to respect expert control and improve usefulness. This case could set important precedents for how AI companies handle personal identity and intellectual property.
So, we’re seeing a moment of reckoning in AI, where innovation meets responsibility. From Nvidia’s huge open model investment to Grammarly’s legal challenges and Zendesk’s strategic acquisition, the AI landscape is evolving fast. It’s clear that respecting expert rights and fostering openness will be key to sustainable AI progress. Thanks for tuning in to The Prompt by Kuro House. Catch you tomorrow for more AI insights.
