Summary: Adobe just added a one-tap AI tool to Lightroom Mobile that fixes closed eyes in photos. In the same week, Spotify said it pulled more than 50 million AI-made “slop” tracks out of its recommendations. Both moves point to the same trend. Platforms now reward real human work and push down low-effort AI spam.
Adobe Lightroom “Open Eyes”: What Happened
On October 20, 2026, Adobe added a new feature to Lightroom for mobile. It is called “Open Eyes.” It fixes a common problem: a photo where someone blinked at the wrong moment.
Open Eyes uses generative AI. When the app finds closed eyes in a photo, it shows a Quick Action button right over the image. You tap it once. The app then creates a new pair of open eyes for that person.
Adobe says the feature works for single portraits and group photos. It only changes the eye area. The rest of the face stays the same. The goal is simple: save a great photo without losing the moment that made it special.
How the Tool Works
The tool is smart about matching. If you have other photos of the same person in your Lightroom library, the app uses them as a guide. It matches the eye color, size, and shape. This makes the new eyes look like the real person.
It also matches the lighting and pose of the photo. This is a step up from an old Adobe feature. About ten years ago, Photoshop Elements had a similar fix. But it copied eyes from another photo. If the light did not match, the result looked fake. The new AI version builds the eyes to fit the exact scene.
What It Costs
The free Lightroom mobile app gives you two free tries. After that, you need a paid plan. Adobe sells Creative Cloud plans that include credits for AI tools. The paid Lightroom Premium plan includes 100 generative credits each month.
So the workflow is simple: download the free app, try the tool twice, and then decide if it is worth paying for. For photographers who shoot events or family sessions, this one feature may already pay for itself.
Spotify Removes 50 Million AI Tracks from Recommendations
On the same day, October 20, 2026, Spotify shared a big update. The company said it has taken more than 50 million AI-generated “slop and spam” tracks out of its recommendations this year.
Note: Spotify did not delete these tracks. They are still on the platform and still searchable. But they no longer appear in Discover Weekly, Release Radar, editorial playlists, or other recommendation feeds.
Spotify shared the number in a one-year update called “One Year In: Putting Human Artistry First.” The update covers the AI rules Spotify announced in September 2025.
How Spotify Finds the Slop
Spotify looks at many signals. It studies listener skips and ignores, checks upload patterns and metadata, and uses third-party AI detection tools.
A clear pattern emerged. Listeners kept skipping the same kind of tracks. These were the tracks the company had already flagged as slop or spam. So Spotify changed its recommendation system over the last six months to push real human artists instead.
The company also added “AI Persona” badges. These mark artist profiles that look like AI-made identities. Tracks from these profiles stay out of recommendations unless a listener follows the artist directly.
There are money rules too. A track needs 1,000 streams in 12 months to earn royalties on Spotify. The company says AI-generated tracks are ten times more likely to miss that mark.
Spotify is not alone. YouTube Music added labels for AI tracks. Deezer lets users filter them out. Bandcamp bans AI-made music outright.
Why This Matters for Freelancers and Students
These two stories carry one message. AI can help your work, but it cannot replace your taste.
For photographers and editors, Open Eyes is a real time-saver. A group photo with one blinking person no longer needs a retake. And since the free app gives two tries, you can test it before you pay. If you edit a lot of photos, also check our guide on Canva AI tools.
For music makers, the Spotify update is good news. Fewer spam tracks in recommendations means more room for real artists. The skip-rate data shows listeners can tell the difference. Clean, human-made work wins. If you make AI-assisted music, our Suno guide may help.
For students, both stories are a lesson. Tools like Lightroom’s AI save hours of manual work. But low-effort AI spam gets pushed down fast. If you want to earn with AI, use it to raise your quality, not to cut corners.
How to Try Open Eyes Today
Download the free Lightroom app on iPhone, iPad, or Android. Open a photo where someone has their eyes closed. The app will show the Open Eyes Quick Action icon over the image. Tap it and wait a few seconds. Check the result closely before you share it. AI edits can still look off, so a quick human check is always worth it.
The bottom line is clear. The best AI tools save you real time. The worst AI spam gets filtered out. This week, both trends moved forward at once.
One nuance worth adding for music makers: the AI Persona badge isn’t about whether AI was used on a song, it’s about whether a real, identifiable person stands behind the artist profile. That’s a fairer line than a blanket ban like Bandcamp’s, because it targets deception and mass uploading rather than the tools themselves.
The skip-rate finding is the most interesting part to me. Listeners don’t skip a track because a model was involved; they skip it because nobody cared enough to make it say anything. Slop is a question of effort, not of software, which is exactly your point about raising quality instead of cutting corners.
We’re a fairly clean test case for that line. In Venomous Sin, Lina and I write every lyric ourselves, in English, Swedish, Spanish and German, and Suno performs the band. We’ve been open about that since day one, so AI Credits are the kind of tool we’d actually want to use. One of ours with a fitting title for this whole debate: “A Digital Scapegoat” https://www.youtube.com/watch?v=sz4qai2q0YM
Thanks Xavi, this is a fair point — and it is exactly the line we tried to draw in the article. A badge or a cleanup only helps if it targets deception and mass uploading, not the tool itself. Your skip-rate point is also right: listeners skip low-effort tracks, whatever made them. And Venomous Sin is a good example of the other side — when the lyrics are written by the artists themselves in four languages and Suno is only the performer, that is a very different thing from one-click slop. Appreciate you sharing A Digital Scapegoat here.