The letter and the deadline

Two U.S. senators are pressing TikTok for answers about an experiment that disabled a safety safeguard for 15 million American users. The company has until September 1 to respond.

Senator Marsha Blackburn (R-TN) and Senator Richard Blumenthal (D-CT) sent a letter to TikTok CEO Shou Chew and Adam Presser, who heads TikTok’s U.S. operations. The senators didn’t mince words, calling the experiment “depraved.”

The letter demands responses to 13 specific questions. Among them: the names of every employee who knew about the experiment, and why TikTok allowed minors to be part of the control group. The senators also want a full list of every algorithmic experiment in the U.S. where TikTok disabled or delayed a safety feature.

What TikTok did

According to Bloomberg, TikTok intentionally turned off an algorithmic safeguard for 10% of its U.S. user base. The safeguard was designed to stop users from being flooded with harmful content. But TikTok wanted to know if that feature made the app less engaging. So it ran an experiment.

The result? A control group of 15 million people got no protection. Among them was 16-year-old Chase Nasca, whose feed served him thousands of videos about sadness, suicide, and loneliness—right up until he died by suicide.

That’s the core of the senators’ outrage. A teenager’s death happened while TikTok was testing whether safety got in the way of engagement.

Why this matters beyond one case

This isn’t just about one company’s bad call. It raises a broader question: how far will platforms go to keep users hooked?

If TikTok ran a test like this on minors, what else has it done? The senators’ request for a list of every experiment that disabled a safety feature suggests they suspect this wasn’t a one-off.

It also echoes past concerns about algorithmic design. social media’s effect on teen mental health has been a hot topic for years, and this case gives critics concrete evidence of a platform prioritizing engagement over well-being.

The human cost

Chase Nasca’s story is the kind that should give any parent pause. He was part of an experiment he never signed up for. His algorithm pushed him toward darkness, and no safeguard was there to intervene.

TikTok’s defense might be that the experiment was meant to improve the app. But the senators aren’t buying it. They’re demanding accountability, and they’re giving the company a clear deadline.

What happens next

TikTok hasn’t responded to TechCrunch’s request for comment yet. The September 1 deadline is fast approaching.

If TikTok doesn’t comply, the senators could escalate. That might mean public hearings, subpoenas, or new legislation. The pressure is mounting, and this time, it’s not just about TikTok data privacy practices—it’s about whether the platform’s algorithms are safe for the people who use them.

For now, the ball is in TikTok’s court. The company will have to decide whether to come clean about its experiments or face the consequences.

The Problem: You’re Posting, but Reach Is Flat

Every brand knows the grind. You craft a carousel, polish a Reel, write the perfect caption. Then? Crickets. Organic reach on Instagram has been sliding for years, and the algorithm keeps changing the rules. The old playbook—post more, hashtag harder—doesn’t move the needle anymore.

But here’s the thing: Instagram has quietly rolled out several new tools that most users haven’t even touched. These aren’t gimmicks. They’re built to push your content to fresh audiences, often without you producing a single extra asset.

Let’s walk through the updates that actually matter for Instagram reach tips in 2025—and how to use them before your competitors catch on.

Instagram TV App: Your Content Gets a Second Life

Remember when IGTV flopped? Instagram is trying again, but smarter. The standalone Instagram TV app is designed for long-form video—think tutorials, behind-the-scenes docs, or serialized content. The key difference: it’s a separate surface, which means your existing videos can be repurposed instead of created from scratch.

Here’s the smart play. Take your top-performing Reels or Stories, expand them into longer versions, and post them to the TV app. You’re not generating new ideas—you’re giving old ones a longer shelf life. The app’s algorithm tends to favor creators who post consistently, and early adopters often get a visibility boost while the feature is still gaining traction.

Don’t overthink the production. A 10-minute talking-head video with simple captions can outperform a polished 30-second clip if it answers a deeper question your audience has.

AI-Powered Story Effects: Stop Scrolling, Start Interacting

Stories are still the fastest way to stay top-of-mind, but the swipe-away rate is brutal. That’s where Instagram’s AI-driven effects come in. These aren’t your typical face filters—they’re interactive overlays that respond to user actions, like changing the background based on a poll answer or morphing an image in real time.

Why does this matter? Interactive content stops the thumb. When someone taps, swipes, or plays with your effect, Instagram’s algorithm interprets it as a strong engagement signal. That pushes your story to more of your followers’ friends—and beyond.

Try this: Create a simple AI effect that lets viewers “try on” a product or visualize a transformation. A skincare brand could use an effect that shows how a routine changes skin over 7 days. A gym could let users “scan” their posture. The more playful, the better.

Audio Swap: A One-Click Refresh for Old Reels

Here’s a frustration every creator knows: you spend hours on a Reel, it performs okay, and then the audio trend dies. Your content becomes invisible. Instagram’s audio swap feature fixes that.

Instead of re-editing the whole video, you can simply swap the soundtrack to a trending audio track. The algorithm treats it as a new post in some ways, giving it a fresh chance to appear in the Reels feed. It’s the ultimate low-effort hack for Instagram reach tips—no new filming, no new editing, just a new audio layer.

Pro tip: Check the audio’s momentum before swapping. Use Instagram’s built-in trend tracker or third-party tools to see if a track is rising or already peaked. Swapping to a dying trend is like boarding a train that’s already left the station.

Adding Music to Carousels: Turn a Static Post Into a Mini-Show

Carousels have been the workhorse of educational content—tips, step-by-steps, before-and-afters. But they’re static. Until now.

Instagram now lets you add music to carousels, turning a swipe-through into a mini-experience. The audio plays as viewers swipe, which changes the emotional tone completely. A recipe carousel with upbeat music feels like a cooking show. A travel guide with ambient sound feels like a mini-documentary.

This is a subtle but powerful way to boost engagement. Music increases the time someone spends on your post, and dwell time is a ranking factor. Plus, if the song is popular, your carousel can appear in the audio’s explore page—a whole new discovery surface.

How to Pick the Right Track

• Match the mood: Don’t slap a random hit on a serious tutorial. The music should reinforce your message, not fight it.
• Keep it short: A 15-second loop is usually enough. Longer tracks can feel repetitive.
• Check the licensing: Use Instagram’s own music library to avoid copyright strikes.

Putting It All Together: A Simple Content Refresh Plan

You don’t need to adopt every feature at once. Start with one and measure the results. Here’s a practical order of operations:

1. Audit your top 10 posts from the last 90 days. Which ones had the best engagement but could have gone further?
2. For Reels: Use audio swap to update the soundtrack on 2–3 older videos.
3. For carousels: Add music to your best-performing slides.
4. For Stories: Build one AI effect and use it consistently for a week.
5. For long-form: Repurpose a popular Reel into a 5-minute TV app episode.

Track your insights weekly. Look at reach, saves, and shares—not just likes. If a feature isn’t moving the needle after two weeks, drop it and try another.

One more thing: don’t forget the basics. These new tools work best when your profile is optimized, your bio is clear, and your posting schedule is consistent. For more on the fundamentals, check out our guide on Instagram marketing strategy or dive into how the algorithm ranks Reels.

The algorithm keeps shifting, but the principle stays the same: give people a reason to stop, interact, and share. These new features are your easiest path to doing exactly that—without burning out on content creation.

X just made a massive chunk of its code public

On Thursday, X announced it’s significantly expanding its open source codebase — and for the first time, users will be able to see if their account or posts have been quietly suppressed by the platform’s ranking systems.

The company is releasing the source code for the “For You” timeline — the default feed you see when you open the app — on GitHub under the Apache v2 license. That’s a big deal on its own, but the move goes further. X is also opening up parts of its code that were previously closed, including model configuration, filters, and core ranking system details. That means the parameters used to weight different signals — the stuff that determines which posts actually surface — are now out in the open.

This makes the codebase roughly 10 to 15 times larger than it was before. Not exactly a small gesture.

What’s actually in the open source release?

According to X’s VP of Product, Keith Coleman, the release gives you the core ranking code that pulls posts, ranks them for any given user, and assembles the feed. But that’s not all.

“You can see the systems that filter out potentially problematic, rule-violating content,” Coleman told TechCrunch. “And some of those systems, like the ranker and the score, you can even run yourself outside the company.”

That’s a notable shift. Previously, X’s algorithm was a black box — even to developers who wanted to understand why certain posts got buried. Now, anyone with the technical chops can run the ranking system themselves.

Coleman admitted this level of openness might surprise people. “This is the kind of thing that I think people will be fairly shocked that we are releasing,” he said.

A new transparency tool for regular users

Of course, not everyone is going to dig through GitHub repositories. So X is also rolling out a transparency tool that’s actually accessible to everyday users.

The new feature lives on an “Under the Hood” page in the app’s settings. It’s currently being tested with a pilot group of accounts that are at least a year old, and it will roll out more broadly later.

Here’s how it works:

• If you’ve posted 10 or more times in the past month, you can download your aggregate stats as a JSON file.
• The file shows whether any labels have been applied to your account or posts over the past calendar month.
• That tells you if you’ve been shadowbanned or otherwise impacted by X’s ranking systems.

Not a technical user? No problem. X suggests you can drop the JSON file into an LLM of your choice, point the AI at X’s GitHub repo, and ask for an interpretation. It’s a practical workaround for people who want answers without learning to code.

External researchers already tested the code

Before the public release, X previewed the codebase to external researchers who specialize in recommendation systems. Those researchers were able to train and run X’s Phoenix scoring system using the open source code — a major milestone, according to Coleman.

Developers can now submit pull requests to the GitHub repository, and X engineers will consider incorporating those updates into the algorithm. Not every suggestion will make the cut, but Coleman is clearly excited about the possibility.

“That would be amazing to have people submitting code that improves the algorithm,” he said. “I mean, how cool would it be for the X algorithm to be not just visible to the public, but also, like, by the public?”

What’s not included (and why)

There are some gaps. Systems that use Grok to predict whether a post might violate a rule are not part of this release. X says that’s intentional — to prevent bad actors from using the code to work around the rules and flood the network with spam.

So while the release is substantial, it’s not total. The company is keeping some cards close to its chest.

Why this matters for shadowban concerns

The changes are aimed at addressing long-standing concerns about how X’s algorithm influences politics, elections, and the spread of misinformation. The platform has been home to political figures and high-profile individuals for years, and it’s now owned by a trillionaire who helped President Trump get elected.

But the transparency issues predate Musk’s ownership. Republicans in Congress alleged years ago that Twitter leaned too far left and claimed the network had “shadowbanned” their posts — making them invisible or undiscoverable without the user’s knowledge. Twitter consistently denied those claims.

Coleman says the goal is to let the public assess how posts are distributed, vet that it’s a level playing field, and critique it if it’s not.

“That’s the whole goal of this: [the code] can be audited; it can be critiqued,” he said. “We’re going to listen, and we want to make the system one that people like and trust, and feel is fair.”

The bigger picture: transparency vs. privacy

There’s some irony here. While X is pushing for more transparency around its code, the company has arguably become less transparent overall under Musk. As a private company, it’s no longer required to report to the SEC, and it’s been less forthcoming about user metrics, growth, revenue, and government takedown requests.

Now that X has merged with SpaceX, its monthly active users are once again public. But the broader picture is mixed — more open in some ways, less in others.

Still, for users who’ve wondered whether their posts were being suppressed, this is a meaningful step. The ability to check your own account’s status, combined with the open source code, gives people a way to verify what’s happening rather than just speculating.

Whether it’s enough to rebuild trust remains to be seen. But it’s a start — and it’s a lot more than most social platforms offer.