Data

Anthropic users face a new choice – opt out or share your chats for AI training 

Anthropic is making some big changes to how it handles user data, requiring all Claude users to decide by September 28 whether they want their conversations used to train AI models. That is a massive update. Previously, users of Anthropic’s consumer products were told that their prompts and conversation outputs would be automatically deleted from Anthropic’s back end within 30 days “unless legally or policy‑required to keep them longer.”

Source: Anthropic users face a new choice – opt out or share your chats for AI training | TechCrunch

Duplicate Works, Diverted Royalties, and What the MLC Isn’t Catching

The Mechanical Licensing Collective was established to streamline digital mechanical royalty payouts.  But even well-intentioned systems can be designed in ways that introduce new risks. That includes the lack of recourse for rights holders and the roll-up of unmatched royalties based on market share. But the biggest issue is the ease with which the system can be manipulated.

Source: Duplicate Works, Diverted Royalties, and What the MLC Isn’t Catching – Digital Music News

Building a Universal Digital Library of All Books Faces Obstacles from Copyright Laws.

The quest to create a universal digital library of all books in the world is a testament to the power of technology and the desire for accessible knowledge. However, this ambitious goal is impeded by stringent copyright laws, which have led to significant legal challenges. Tech companies like Google and AI firms such as Anthropic have developed extensive digital text collections, some of which were sourced from shadow libraries like LibGen and Sci-Hub.

Source: Building a Universal Digital Library of All Books in the World Faces Obstacles from Copyright Laws.

French press take on digital databases to defend journalist copyright against AI

Two professional organisations representing 800 newspapers and magazines employing over half of journalists in France announced Monday that they are taking “coordinated action” against public datasets used to train generative artificial intelligence services, such as ChatGPT. The Apig, the general news medial alliance, and the Sepm, the magazine publisher’s union, aim to remove their members’ content from Common Crawl, C4 and Oscar.

Source: French press take on digital databases to defend journalist copyright against AI

AI giants race to scoop up elusive real-world data

Over the past two months, OpenAI has tied up with e-commerce majors Shopee and Shopify, while Google and Perplexity have doled out free access to their advanced AI tools to some users in India. Experts believe these moves will help the companies access structured consumer queries, product behaviors, and transactional data — training signals that are often unavailable via public data alone.

Source: AI giants race to scoop up elusive real-world data

Fastly warns AI bots can hit sites 39K times per minute

Cloud services giant Fastly has released a report claiming AI crawlers are putting a heavy load on the open web, slurping up sites at a rate that accounts for 80 percent of all AI bot traffic, with the remaining 20 percent used by AI fetchers. Bots and fetchers can hit websites hard, demanding data from a single site in thousands of requests per minute.

Source: Fastly warns AI bots can hit sites 39K times per minute

Perplexity's Comet AI Web Browser Had a Major Security Vulnerability

Comet, Perplexity’s new AI-powered web browser, recently suffered from a significant security vulnerability, according to a blog post last week from Brave, a competing web browser company. The vulnerability has since been fixed, but it points to the challenges of incorporating large language models into web browsers.

Source: Perplexity’s Comet AI Web Browser Had a Major Security Vulnerability

More UK news publishers are adopting ‘consent or pay’ advertising model

Sixteen of the 50 biggest news websites in the UK are now using a “consent or pay” model to allow users to pay to reject personalised advertising or even avoid ads altogether. UK publishers began to implement the model last year as the Information Commissioner’s Office cracked down on the requirement for the biggest sites to display a “reject all cookies” button as prominently as the option to “accept all”.

Source: More UK news publishers are adopting ‘consent or pay’ advertising model

AI Becomes Hollywood’s New Secret Weapon in Talent Negotiations: ‘The Data Doesn’t Lie’

Where many in Hollywood view AI as something of a bogeyman, fearing widespread job loss, others see opportunity. Acharia, who hails from a tech background as an entrepreneur before pivoting to management, is one of a small but growing legion of talent reps who are harnessing AI data to better advocate for and negotiate on behalf of their clients on future deals.

Source: AI Becomes Hollywood’s New Secret Weapon in Talent Negotiations: ‘The Data Doesn’t Lie’

AI crawler Firecrawl raises $14.5M, is still looking to hire agents as employees

The Firecrawl founders are working on tools to help website owners, publishers, and other content creators “get paid when AI uses their content. We think this is the way it should be,” CEO Caleb Peffer said. While there have been lots of efforts around this idea from big names like Adobe and Getty, Peffer feels that Firecrawl has an edge because it’s already working with those who are scraping data.

Source: AI crawler Firecrawl raises $14.5M, is still looking to hire agents as employees | TechCrunch

Get the latest RightsTech news and analysis delivered directly in your inbox every week
We respect your privacy.