Free Playbook: Dodge the AI content trap

Plus: Instagram's big AI backlash, the EU's threat to Meta, and the rest of this week's must know social media updates.  ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌ ‌

Read online Why the UK's teen social media curfew is a 'dog's dinner' + The 5 social strategy mistakes killing engagement in 2026 Best Geekout Experience Geekout looks better in your browser. Open in Browser → Advertise in Geekout Reach 30,000 social media pros 60%+ Open Rate | 2-4% Ad CTR Previous sponsors: VEED, Sprout Social, TikTok Check Availability → Hello Geeks The UK government has declared war on teen doomscrolling, and armed itself with a toggle. This week's big idea: a social media curfew for 16 and 17-year-olds. Midnight to 6am, apps like Instagram, TikTok and YouTube off by default. Autoplay and infinite scroll off too. Sounds bold. Then you read the small print. 🤦 It's a smoke alarm with a snooze button. Ministers called social media "addictive" in their own announcement. The EU says the same. A US jury just agreed. Then they built a barrier a teen can switch off in about four seconds. It's NOT a curfew. It's a mildly annoying settings prompt dressed up as safety policy. Ellen Roome, whose 14-year-old son Jools Sweeney died in 2022 — a death she believes was linked to an online challenge — put it best: it's like "offering a 17-year-old a bottle of alcohol," left just out of reach, that "they can just drag it back in." Shadow UK education secretary Laura Trott called it a "dog's dinner". She's not wrong. - - Ministers also confirmed there'll be no restrictions on VPNs, pointing to research that says only 7 to 10% of teens use them. Cute. But make something forbidden and watch that number climb: when Australia's ban went live, one VPN saw installs jumped 400% in a day. The real fix is boring: check age once at the door, not a hundred times at every app. That means leaning on Apple and Google, not a public scrap with Meta. And for parents who've lost children, with deaths linked to their social media use, the appearance of letting Meta off the hook is, understandably, a bitter pill to swallow. In trying to appease everyone, the government's annoyed everyone, and done next to nothing to actually keep teens safer online. Sky News invited me onto its flagship evening news programme to give my (very calm, may be slightly ranty) take on it 👇 - — Matt __________ 3 cool things… 🎯 The 5 mistakes quietly killing most social strategies → Number 2 is the one I see constantly → [Grab the free playbook] [ad] - 🚀 See blowing up on YouTube before it's obvious → This tool tracks millions of videos with 30-min rank snapshots → [See this week's breakouts] - 🤪 Love memes and weird viral videos?→ You need this newsletter. Ryan Broderick's Garbage Day is genuinely one of my favourites → [Check it out] SPONSORED ▶︎ FREE PLAYBOOK 5 Mistakes Quietly Killing Your Social Strategy 💀 - Here's an uncomfortable one… Could a competitor publish your last 10 posts unchanged, just swap the logo? If you're not sure, that's your answer. AI didn't create the content noise problem. It just handed it a megaphone 😬  82% of teams now use AI for content, but only 32% have a clear audience to point it at. More output, less impact. This FREE playbook from Agorapulse breaks down the 5 mistakes that are silently killing your social strategy. And how to fix them: 🎯 Publishing without intent — the 4-question scorecard that kills a weak post before it wastes a slot 📊 Confusing volume with strategy — why 93% of pros already know posting more is a trap, and what to measure instead 🤖 Feeding AI the wrong context — the exact "give it this, not that" list that stops your content sounding like everyone else's ✏️ Skipping the editorial filter — the "killer test" one question that separates a real post from noise 🔁 Scaling what doesn't work — how to spot a lucky spike before you pour budget into repeating it Plus a 30-minute strategy audit and 4 ready-to-use templates. Steal all 5 fixes, before your next content review → Get the Playbook Send me the Playbook → Instagram and Facebook will likely require a redesign after EU rules they’re ‘addictive’ The addictive design threatens the ‘physical and mental wellbeing of users.’ The Verge • Dominic Preston Meta removes Instagram’s AI remix option after backlash [UN-MIXING] X tweaks its algorithm to make the feed feel less hostile [SANITIZING] TikTok faces child safety probe in the UK over age verification failures [RULE-BREAKING] UK government proposes social media curfew for teens aged 16 and 17 [RESTRICTING] GEEKOUT FAV ▶︎ TOP NEWSLETTER The newsletter I read to make sense of it all 🗑️ - Every week in Geekout I tell you what the platforms shipped… Garbage Day tells you why the internet feels the way it does underneath all of it. It's the award-winning newsletter from tech reporter Ryan Broderick. Three times a week (give or take), he crams as much internet junk as possible into one email so you don't have to go hunting for it. Memes, weird viral videos, internet drama, whatever the heck is happening with AI, it's all there. If your job is social, this is the context that makes the news land. It's genuinely one of my top 5 newsletters, and one of the few I read start to finish. Make your inbox a little less professional and remember what it was like to have fun online again. ↪︎ Check out Garbage Day, it's free 🗑️ Get Garbage Day X gets an actually very good birthday present X just tweaked its algorithm to make it more friendly, less battleground TechCrunch • Lucas Ropek Since he took over Twitter and renamed it X, we’ve got used to Elon Musk making the platform worse. In the era of Elon, we’ve seen a ramping up division, not least by prominently promoting his own political views, trying to shape Grok in his own image, practically shrugging his shoulders at non-consensual nude AI images, and taking a nonchalant attitude to any government wanting to regulate X. But now it seems the platform has done something against the grain and genuinely good for its humanity. X has tweaked its algorithm to make the feed feel less hostile by showing more mutual follows. Instead of seeing content that makes your blood boil, you’ll see more of the stuff you’re actually interested in. As TechCrunch notes, “the change may not drastically revamp the site’s user experience, but may make X feel a little bit more like a community rather than a torrent of disparate voices shouting into the digital abyss.” And that, for me, has been the core problem on X for too long. It has long incentivised sharing extreme views, because they’re what get people talking, and thus get your posts to actually show up in the users’ feeds. The result is that the politicians and journalists who still for some reason use X as a barometer of public opinion believe fringe political views are more popular than they are. The platform’s ability to shift the Overton Window and make more extreme ideas more acceptable must surely have been part of Elon’s reason for buying it. He has long presented his own beliefs as being critical to the future of humanity. But at some point logging onto an argument machine every day stops being fun for users. In the week that Twitter/X turned 20 years old, it’s good that the platform might be getting an little of its old utility back. …but Elon being Elon, I wouldn’t bet on it staying that way. That guy just loves the drama too much. Meta   /  Facebook  '[Image: Meta] Meta AI will now alert parents about teens’ suicide or self-harm chats and is developing emergency-service alerts for users at imminent risk ...Also, Meta says teen chats will default to 13+ safety settings, blocking sexual, romantic and other sensitive responses. Instagram and Facebook have expanded AI translation for Reels into more languages Meta’s new invisible AI image watermark becomes much less effective if an image is heavily cropped, tests show Meta has asked Apple to let its devices pair across a user’s iPhones and iPads, like Apple’s own AirPods do Meta has been accused of using biased AI targeting for mass layoffs. The company denies this __________ Instagram  [Image: Meta] Instagram has added timestamps for when people viewed your Stories. But only if you’re an Instagram Plus subscriber Instagram looks set to charge for AI features in the future as Adam Mosseri says they’re “very expensive” to run Instagram now lets you add an AI creator label to your profile Instagram has added new ‘Fill in the Blank’ profile banner Instagram Edits is getting font importing, song lyrics, plus video effect search on iOS __________ Threads   /   WhatsApp  [Image: Meta] Threads has introduced flair tags to make Communities more useful Threads is teasing a redesigned layout for the Activity tab ChatGPT is back on WhatsApp in the EU after the European Commission ordered Meta to reinstate competitors for free WhatsApp is working on a new Android homescreen widget for recording voice notes WhatsApp is developing birthday reminders based on dates already saved in your phone’s contacts WhatsApp is building its own encrypted cloud backups for iOS, as an alternative to iCloud. Up to 2GB would be free, with 50GB for 99 cents per month WhatsApp is rolling out support for a second account in the Android Business beta app WhatsApp Channels could soon let admins pin key updates WhatsApp is rolling out notifications when someone pins a message in a chat or group WhatsApp usernames are starting to become useable, with some users being able to contact new people + businesses with sharing their number __________ X / Twitter  [Image: X] X now lets brands boost posts that mention them The EU has given Elon Musk’s X six months to prove it can be more transparent and follow its digital rules Twitter / X turned 20 years old this week. And now I feel old xAI is suing a Grok user over alleged CSAM deepfakes, in a test of who is responsible for what AI makes Elon Musk says X will open-source its entire codebase X is testing a new profile UI with different CTA buttons: Mention and Message __________ TikTok  TikTok has quietly revamped its failed Instagram rival app Whee as ‘TikTok Chat’, shifting the focus from photo sharing to messaging TikTok is rolling out a new pilot program where it will run most Shop operations for some sellers in the U.S., if they pay $10,000 + commission fees TikTok signed a global content deal with the NBA and WNBA, combining game highlights, creator access, live-game discovery and new advertiser opportunities TikTok has defended its child safety controls as pressure from the EU builds An inquest has been reopened into the death of 14-year-old boy after a ‘TikTok blackout challenge’ TikTok Shop now shows UK-based sellers exactly what drives product sales TikTok Shop is banning AI voices from shopping livestreams TikTok now lets artists upload original music directly within the app TikTok has been spotted offering ‘Songs of LIVE’ a gift feature for creators __________ LinkedIn   /  Reddit    /  YouTube  [Image: Beehiiv AI] LinkedIn has been pressed to do more to tackle online stalking after a jailed offender’s account has remained online Reddit is starting to question whether its AI data deals hurt its ads business Google Image Search is becoming a Pinterest-style discovery feed Telegram has added a rich text editor, communities, ephemeral messages in groups, and improved GIF search YouTube has appealed a landmark Los Angeles verdict over alleged social media addiction harms. YouTube now lets creators in its Partner Program package video playlists as shows with seasons and episodes YouTube is rolling out its ‘Ask YouTube’ AI search tool more widely on the web in the US. __________ Also spotted this week… Meta is opening up its ads MCP server to all developers, making its ads tools easier for AI agents to use Meta will start charging businesses for its AI agent across WhatsApp, Instagram and Messenger from August 1. X Ads are getting Grok integration for creating and managing campaigns X is using Grok to crack down on engagement bait and stolen content, redirecting more than $1m to original creators Google now lets you star in your own AI videos. Google is adding custom digital avatars to Vids, built from a user’s selfie and voice recording YouTube is testing an AI-powered Studio Insights revamp with smarter performance, content and audience loyalty data YouTube clarified which which AI slop it will de-monetise as it renames its ‘Inauthentic Content’ policy to ‘Generic and Repetitive Content’ Agentio is bringing AI-powered creator ad matching to Meta. Brands can now use its platform to plan and launch creator ads across Facebook and Instagram Earlier this week, I explained to BBC News how audiences respond to rivalry without hostility, after England’s Jude Bellingham and Norway’s Erling Haaland became the World Cup's top bromance. - -- __________ - - __________ - Victoria Beckham managed to become a World Cup sensation this week… by doing nothing. Her (non-)response to an England goal went viral… - - __________ - - But at least some people could relate to her behaviour…! - - __________ - How ‘unbothered’ Victoria Beckham became one of the World Cup’s biggest memes the Guardian • Guardian Staff __________ Also viral this week: How Erling Haaland became the most viral player of the World Cup Ahead of Norway vs England, Norwegian Air publicly challenged British Airways to a one-day Instagram logo swap. The first post took around 30 minutes from idea to publishing. Rachel Karten has taken a deep dive in her newsletter. Some fun insights: Norwegian pre-made both winning and losing posts so it could react instantly after the match. The campaign generated 25m views, 1.5m+ reactions and 55K comments. Norwegian later turned the attention into a promotion, which it said performed well Read the full story 👇 My favorite World Cup social moment Rachel Karten __________ Advertisers say Meta’s AI ad tools are creating brand chaos… Advertisers and agency executives told Business Insider that AI errors are becoming routine. Reported problems include altered products, garbled text, twisted limbs and unwanted creative changes. Some advertisers claim bugs have switched AI enhancements on without approval 😮 Meta says its tools can make mistakes and advertisers must review every output. Well yeah, but at some point, you might as well just have humans more involved from the start! Meta's AI advertising dreams have become a nightmare for brands Business Insider __________ More brand news and views: How brands like Doordash and Coors Light are embedding themselves in World Cup hype on Reddit Unilever is using AI to manage campaigns with 300,000 creators Why FIFA and YouTube are betting big on creators this World Cup "…It is not that hard to build a token incinerator…"   – Instagram boss Adam Mosseri on Meta cutting back on its “silly” AI overspend __________ Also quoted this week: “I don’t think we should filter out AI content… I think we should let you know if content is AI content or not.” - Instagram boss Adam Mosseri thinks users should be in control of whether their feed is full of AI content or not __________ “Our intent was to provide a useful creative tool and to give people control over whether their public content could be referenced in this way. We've heard the feedback that this feature missed the mark, so it's no longer available." - Meta explains its removal of the tool I called “creepy” in last week’s Geekout Reelful uses AI to turn your camera-roll clips into ready-made social videos Just describe the story, upload photos and videos, then record a 30-second voice sample. The app writes a script, adds captions, music and voiceover, then assembles the edit. It can also animate still photos into AI-generated video clips. You can then refine the result by chatting with Reelful to change scripts, soundtracks or scenes Reelful's AI turns your camera roll into short-form videos for social media TechCrunch • Aisha Malik __________ Also this week: HyperTexting lets you follow websites, blogs, newsletters and podcasts in a scrollable feed, turning the open web into a social-like experience TikTok users are framing the diarrhoea parasite Cyclospora outbreak in the US as a way to get “skinny for summer” 🫪 The CDC has reported 1,645 confirmed cases across 34 states since May 1, with 141 hospitalisations. Videos joking about “Cyclospora skinny” and “the diarrhea diet” have spread through SkinnyTok-style communities. Health experts warn any weight loss is mostly water and can come with severe dehydration and lasting gut issues. TikTok Is Turning a Diarrhea-Causing Parasite Into a Weight Loss Hack: ‘Risking It All for the Cyclospora’ VICE • Ashley Fike __________ More WTF this week: South African pro cyclist Willie Smit was removed from a race for wearing Meta’s smart glasses AI-generated influencers are flooding TikTok Shop Fitness influencers are being linked to an illegal steroid market on Telegram Telegram’s shortlinks domain was taken offline for a day. Its likely a single link using the shortener was in violation of US sanctions Plastic surgery influencers are becoming looksmaxxing watchdogs YouTube shared pro-tips on how to use Shorts to drive long-form channel growth… How to convert YouTube Shorts views into long-form channel growth blog.youtube __________ Also this week: Sharing another creator's Reel to your story can help increase their reach on Instagram, but it is not the primary way creators gain traction, Adam Mosseri says Instagram Reels do not have a set "shelf life" for recommendations, Adam Mosseri says. While they are most frequently recommended in the first day or two after being posted, they can continue to be recommended long after that. 25+ MILLION   – The number of global subscribers to Snapchat+ and Snap’s other subscription plans in 2026 __________ More social stats: IN VOGUE: Madonna’s live video event for ‘Confessions II’ set a new TikTok record with 2.1 million global viewers QUIET TIME: 55% of Americans say they post less on social media Also: 53% are also more selective about who sees their posts, and 47% have deleted a social or messaging app due to stress or anxiety HARMFUL: Eating disorder videos now appear in one in nine teen-account recommendations on YouTube. CREATOR ECONOMY: YouTube says creators added $60 billion to the US economy and supported 540,000+ jobs in 2025. When did social media stop being fun? Maintaining an online presence feels like work for 51% of Americans, rising to 60% among Gen Z. - __________ Also charted this week: INFLUENCED: Who gets more health advice from influencers: men or women? WHO’S ONLINE: Discord’s US unique visitors by age Music influencers are getting paid to promote songs — and they're not telling you Social media can make a song go viral — even push it up the pop charts — seemingly overnight. Perhaps it's not surprising that some influencers are being paid to promote music without disclosing it. NPR __________ Also spotted this week… Meta’s Oversight Board says leading AI models might be restricting free expression as it tries to extend the company it emerged from. [Engadget] Google is using social media signals to mask AI search click loss. [Search Engine Journal] Hacked Suno files show the AI music startup scraped audio from YouTube Music, Deezer and Genius. [404 Media] UK regulator Ofcom wants social media platforms to stamp out scam ads or face big fines. [The Independent] …and Ofcom says age checks are now being used at an “unprecedented scale” across social, dating, gaming sites and more. [Mashable] The European Commission is set to propose age-appropriate social media restrictions for the EU after the summer. [The Verge] …but the recommendations in a new EU report are tougher than they look. [Politico] Bluesky has made Toni Schneider its permanent CEO. [Engadget] Meta is expanding its Hyperion data centre from 2GW to 5GW capacity. [Quartz] Sprout Social is cutting 20% of its workforce. [MarketWatch] Meta plans to hire a top Amazon exec as it weighs spinning up its own cloud computing business [WSJ] Disney is using a glitzy L.A. premiere to court creators with Hollywood ambitions [TubeFilter] Trump Media plans to sell instant access to 'market-moving' social posts by launching a Truth Social API [BBC News] New features and tests: Snapchat has announced a new partnership with Dick’s Sporting Goods to provide direct data insights into how Snapchat campaigns drive sales activity in Dick’s stores across the US. [Social Media Today] Hinge users can now get friends to hype them up on their profiles. [Engadget] TikTok appears to be testing Apple Pay integration. [@jonahmanzano] Around 400 attendees gathered in Maryland for three days of memes, philosophy and organised weirdness at Vibecamp. The event grew from ‘This Part of Twitter’, an undefined community shaped by Rationalism, AI and internet culture. Activities at the event included Twitter roleplay, AI fiction, metaethics debates and prolonged eye contact 😳 Welcome to Vibecamp, an IRL Gathering for the Terminally Online Rolling Stone • Nick H. Penniman __________ Also worth a read: From Twitter to X: 20 Years of Memes, Movements and Hot Takes How Meta has spent weeks sowing confusion over the existence of its face recognition tech Will the social media ban be the death knell of teen style subculture or its revival? ‘Surveillance Chic’: can Meta make its smart glasses a viable fashion accessory? 20 things social media algorithms do to your brain that you didn't agree to Your next customer may discover your brand on TikTok before Google Today is Emojipedia’s World Emoji Day! 🥳  To mark the event, here’s a look at 9 new(ish) emojis coming soon to your phone. Unicode is adding a meteor, lighthouse, pickle, eraser, net and cracked smiling face emojis. New left and right thumb signs are also planned. The update also separates comets from meteors, cucumbers from pickles and butterfly colours. They are set to launch in the spring of 2027. A sneak peek at nine new(ish) emoji on the way - 9to5Mac 9to5Mac • Ben Lovejoy - Meanwhile, congratulations to Distorted Face 🫪, the most popular new emoji of 2026! That's everything worth knowing about this week. I'm off to rewatch my Sky News hit to check if I actually said “world-leading online safety” and "with a settings prompt," in the same sentence… — with a straight face. 😆 - Gif by theoffice on Giphy - Goodbye, geeks! - P.S. This newsletter is edited by Martin SFP Bryant     © 2026 Matt Navarra Media Ltd 7 Rhos Ddu Penarth, Vale of Glamorgan CF64 3HS, Wales Unsubscribe Terms of Service