New social media features and updates to know this week
Including updates from Meta, TikTok, Pinterest and more.
It’s a lighter week for social platform updates, but the changes that did land are worth noting.
Meta’s boosting its AI assistant while Instagram has already made upgrades to its new editing app. Meanwhile, TikTok’s now turning third-party DMs into lead-gen machines, Pinterest is dialing up user control on AI images and YouTube’s blurring the lines – literally – for more risqué search results.
Let’s get into it.
Meta
Meta has upgraded its AI assistant on WhatsApp, Instagram, Facebook and Messenger, introducing faster responses, more natural voice interactions and better personalization using its new Llama 4 model.
The company made the changes alongside the launch of its standalone Meta AI app.
Now, the assistant responds in a more conversational tone, supports image generation and editing, and offers answers tailored to each user’s activity across Meta platforms. Users who link their Facebook and Instagram accounts through the Accounts Center receive tailored suggestions.
Meta also introduced a Discover feed that highlights how people are using the assistant, including shared prompts. The assistant can remember user preferences and maintain context in ongoing conversations.
While the standalone app includes extra features like document generation and integration with Ray-Ban Meta glasses, Meta has brought most of the enhancements directly into the social and messaging platforms where users already have a familiarity with the AI tool.
Meta also appears to be working on the cross-promotion of its platforms, as evidenced by the testing of a new prompt on Threads that encourages users to share their posts directly to Instagram Stories.
As social media commentator Matt Navarra shared, some Threads users are being prompted to cross-post their content to Stories to “get more eyes on your posts.” Meta already allows for Facebook-to-Instagram sharing and is expanding cross-posting features on WhatsApp.
While sharing Threads posts to Stories may increase visibility, it’s important to note that not all content works well across platforms. Just because you can cross-post doesn’t always mean you should.
Despite that, the push to cross-post from Threads to Instagram is aimed at driving engagement and bringing more attention to both platforms. With Threads surpassing 350 million monthly users, Meta hopes this feature will encourage users to engage with both platforms more frequently.
On the Instagram-specific front, the app has launched the first updates to its new Edits video editing app, growing its font options to 125, increasing voice effects to 15 and installing a simplified editing process. The new “Apply all” button lets users apply filters, effects and transitions across all clips at once.
The platform, which is similar to TikTok’s CapCut, also appears to be in the middle of testing a new feature that lets users further customize their Notes. App researcher @varma.adwaith posted a screen grab of an IG prompt that reads: “Now you can customize your note by adding color and an emoji. Anyone can choose to use your note style as their own and your name won’t appear next to it.”
The business-centric platform has also updated its identity verification process to support users who don’t use their legal names on the platform.
LinkedIn users can now add a legal name as a secondary identifier during verification. The name appears in parentheses next to their display name – visible only to logged-in users – and helps confirm identity through government-issued ID. Public profiles and search results will still show only the chosen display name.
While the option to list a secondary name isn’t new, LinkedIn now allows users to use it during the verification process. Removing the legal name will also remove the verified badge.
Pinterest has added labels to content it detects as AI-generated and is introducing new controls to help users manage how much of that content they see.
After months of testing, the platform will now display an “AI modified” label on image pins that appear to have been created or altered using AI.
Detection relies on both metadata and machine learning classifiers, which Pinterest says will improve over time. A formal appeals process is in place for creators who believe Pinterest has mislabeled their content.
Pinterest is also experimenting with a “see fewer” option on AI-generated pins in categories like beauty and art, allowing users to signal they want less of that type of content. The option is available in the three-dot menu and will expand to additional categories.
Matt Madrigal, Pinterest’s chief technology officer, described the tools as a way to empower users to “make more informed choices about the content they see.”
The company plans to introduce more of these tools in the months ahead.
TikTok
TikTok has added a new lead generation option to its Promote ads, allowing brands to drive direct messages to third-party apps such as WhatsApp and Facebook Messenger as well as popular Asian messengers like LINE and Zalo.
This new feature allows messages from potential customers to be sent directly to the TikTok inbox, now known as “More TikTok messages.”
Previously, Promote campaigns could only direct users to TikTok’s native inbox. With this update, advertisers can now select an external messaging app as their call-to-action, targeting users who have that app installed.
This new option will give brands more flexibility in how they collect leads, especially as direct messaging grows in popularity for customer engagement.
To set up a third-party Promote ad, choose “Get Leads” as the goal in the TikTok app and then select “Other messaging apps.” Pick one messaging platform, add the correct contact link, choose the video or photo creatives, target and audience, set a budget and duration and then confirm payment.
Brands can only target users who are 18 or older.
YouTube
YouTube has started testing blurred thumbnails in search results for queries that often surface sexually themed content.
The platform blurs video thumbnails while keeping titles, channel names and descriptions fully visible. While the videos meet Community Guidelines, YouTube says it implemented the test because some viewers may prefer to avoid them.
Viewers can choose to unblur any thumbnail if they want to preview the image.
YouTube limited the test to a small group of users and wants to see whether the feature helps people avoid accidentally viewing sensitive content.
Casey Weldon is a reporter for PR Daily. Follow him on LinkedIn.
