Shopping features on Twitter pose risks to content moderation and could be used “in ways that lead to individual or societal harm,” according to an internal memo sent to a team of Twitter employees.
The memo, part of which was obtained by: The edge, was sent to a group of Twitter employees in different teams in early July, according to one person in the know. In the email, some existing and unreleased Twitter shopping features are categorized as high risk, with the email warning that content moderation is not a priority when shopping.
Twitter Shopping was first introduced last summer and allows brands to list items for sale and pin a handful of products to the top of a merchant’s profile. Unlike similar features on Instagram, users cannot purchase the product directly on Twitter – items for sale link to a merchant’s website instead. An expanded version of the store module was introduced earlier this year, allowing merchants to place and showcase up to 50 products in their storefront, and was available to all US merchants beginning in June.
A section of the memo titled “risk assessment” categorizes several elements of Twitter’s ecommerce tool as “high.” A high-risk item is merchant-generated fields such as store names and descriptions, which the memo warns could be used in malicious ways by malicious parties.
Twitter’s shopping feature allows anyone with a professional account that sells items in the US to manually add products for sale to their profile. When selecting items to display in the expanded Twitter store, sellers can add a custom store name and descriptions directly on the Twitter dashboard. It is in these areas that people who work on shopping say there are risks.
The memo paints a picture of a simple process to find and remove potentially offensive or harmful content from Twitter shopping. The platform does not yet have a policy on what is considered a store name or description violation, according to the memo, and it lacks guidelines on how to handle violations. The memo also notes that Twitter lacks the tools to detect violations in store names and descriptions, and there is no way for users to report storefronts for content in these fields.
A key selling point of Twitter’s shopping features is shareability, and the company has introduced other updates, such as reminders that allow customers get a notification and tweet about brand new releases. Currently, Twitter users have the ability to click into merchant stores, view products, and click through to merchant websites, but window sharing is not yet available. The internal memo lists the window-sharing capability as “high” in the risk rating, saying that if the feature is released, it could lead to further amplification of malicious content, increasing the visibility of content that violates Twitter rules. increases.
Twitter has some automated detection mechanisms for individual products for sale, according to the memo. But proactive measures to detect violations are “limited,” and the company has minimal staff and resources for further review.
“Shareable stores therefore increase the likelihood that users will see stores that infringe stores or goods that infringe a store,” the memo reads. “It can also incentivize bad actors to amplify harmful or violent goods by tweeting and sharing their own shop.”
In a statement to The edge, Twitter spokesman Lauren Alexander confirmed the memo’s authenticity and said it was part of a new feature review led by the product trust team. Alexander says the reviews are designed to allow different teams to provide input to ensure new product releases are safe.
“We are always working to improve the security of our service, especially with the addition of new products and features,” says Alexander. “We have deliberately taken the time to test our new retail surfaces so that — before scaling or expanding into new markets — we have the opportunity to gather lessons to better inform our health and safety approach and address additional infrastructure and policy needs. develop. .”
In recent years, Twitter has tried to diversify its revenue streams beyond advertising by experimenting with a paid subscription service, live broadcasting on Spaces with tickets, and paid Super Follows. While the platform has rolled out and expanded its trading tools, it is lagging behind other social networks such as Instagram and TikTok; in his attempted takeover of Twitter, Elon Musk indicate that payments had the potential to make up a significant portion of the business.
Earlier this week, Peiter “Mudge” Zatko, Twitter’s former head of security, said the company suffered from negligent security and bot removal vulnerabilities, among other things, in a comprehensive whistleblower report. Shortly after the notification of the complaint, Reuters reported that the platform would combine a “health” team working to reduce misinformation and other toxic content with a service team working on spam accounts.