Skip to main content

Wordpress.com is selling user data to AI companies









In September 2023, WordPress.com quietly updated a developer page about accessing a "Firehose" of around a million daily WordPress posts, clarifying that these feeds are intended for search engines, AI products, and market intelligence providers. This change, which was not previously noted, has sparked discussion, especially after revelations that Automattic, the owner of WordPress.com and Tumblr, plans to share user data with OpenAI and Midjourney.


The recent buzz surrounds questions about which WordPress blogs are included, data retroactivity for opt-outs, and the lack of transparency from Automattic. This company has been selling access to post data for years, including through the Jetpack plugin, but has recently announced exclusions for select AI partners.


Automattic's deals with OpenAI and Midjourney are particularly contentious as they aim to enhance generative AI tools. This data sharing, facilitated by partners like SocialGist and DataStreamer, offers insights into market trends and user behavior, but raises concerns about privacy and data usage.


SocialGist, a major player in data aggregation, emphasizes its access to WordPress and Tumblr posts, catering to market research and AI training. Despite assurances from Automattic about data usage, questions remain unanswered about terms of use, privacy features, and enforcement mechanisms.


The complex data supply chain underscores the challenges of tracking data usage and enforcing policies. Users have limited visibility into how their content is shared and utilized by third parties, raising ethical and privacy concerns.


As the landscape of data sharing evolves, users are left grappling with the implications of their digital footprint being used for purposes beyond their control. The lack of transparency and accountability in the data economy highlights the need for greater regulation and user empowerment.

Popular posts from this blog

Signal Introduces Usernames for Encrypted Messaging: A Secure Way to Connect

Signal, the encrypted messaging service, is launching a new feature in the coming weeks: support for usernames. This beta feature allows users to establish unique usernames, enabling connections without divulging phone numbers. source: Signal Blog To create a username, navigate to your settings and select "Profile." Once you've chosen a unique username, generate a QR code or link to share with others. Recipients can connect by entering your username into the chat bar. Usernames can be changed at any time, though previous usernames may be claimed by others. Signal began testing usernames last fall. Unlike social media platforms, Signal usernames do not serve as logins or public handles. They offer a discreet means of communication without revealing personal phone numbers. While a phone number is required to register for Signal, sharing it is optional. Usernames remain private and do not appear on profiles or in chats unless shared explicitly. As Randall Sarafa, Signal'

AT&T Resets Millions of Customer Passcodes After Data Leak: What You Need to Know

AT&T recently confirmed a significant data breach affecting over 7.6 million current customers and 65 million former customers. The leaked information, which dates back to 2019 or earlier, includes personal details like names, addresses, phone numbers, and social security numbers. Fortunately, financial information and call history were not compromised. In response to the breach, AT&T has reset passcodes for affected customers. Passcodes, usually four-digit numbers, serve as an additional layer of security when accessing accounts. However, security experts warn that the encrypted passcodes leaked alongside customer information could be easily deciphered, posing a risk of unauthorized account access. Affected customers are advised to set up free fraud alerts with major credit bureaus and remain vigilant for any suspicious activity related to their accounts. AT&T is proactively reaching out to impacted customers via email or letter to inform them about the breach and the meas

Safeguarding Internet Privacy: Supreme Court of Canada Upholds Protection of IP Addresses

In a recent ruling, the Supreme Court of Canada affirmed the significance of privacy rights concerning internet addresses. The court declared that police cannot simply obtain a suspect’s IP address without a court order, emphasizing the expectation of privacy that Canadian residents hold for such information. The court's decision stemmed from a case in 2017 involving Calgary police investigating fraudulent online activities at a liquor store. Initially, police demanded IP addresses from a credit card processor, which eventually led to obtaining subscriber information from Telus. This information was pivotal in making arrests and securing convictions in multiple offenses. Despite previous convictions, the accused contested the legality of obtaining IP addresses without proper authorization. The Supreme Court, in a 5-4 decision, asserted that IP addresses carry a reasonable expectation of privacy, necessitating judicial approval before access. The ruling emphasizes that obtaining jud