Daily Marketing News, September 28, 2026: TikTok caps teen screen time in Alabama

TikTok will pay the state of Alabama $100 million and limit teens to 2 hours a day in the app. On the same day, X and Instagram both had news about AI agents that work on the user's behalf. Instagram showed how such an agent writes content plans for creators, and X started paying people who build ready-made workflows for agents.
Story of the day: TikTok limits teens in Alabama to 2 hours a day in the app
On September 27, 2026, Social Media Today reported that TikTok agreed to new rules for teens in the US state of Alabama. The state had sued TikTok over the app's harm to teenagers. The rules are part of a $100 million settlement, which kept the case from going to trial. TikTok was accused of allegedly designing addictive features on purpose, including the feed algorithm that keeps people from leaving the app.
TikTok made 3 commitments. Teens get no more than 2 hours a day. From midnight to 6 a.m. the app is closed to them. Notifications are switched off during school hours. Until now, only Meta had a similar set of restrictions: the same terms were part of its recent $18 billion settlement with California. TikTok's list of commitments says nothing about the feed algorithm itself. What's being limited is the hours during which the feed can hold on to a teenager.
This matters beyond Alabama. TikTok and Meta face similar lawsuits in at least 27 more US states. If these rules spread nationwide, brands and creators with a teen audience in the US will have less time to reach it. 6 night hours will drop out entirely, and notifications about new videos will stop arriving during school hours.
What to do right now. Open your TikTok account analytics and check your audience's age and country. If a noticeable share is US teens, move your posts to after school and before midnight. Content that relies on instant notifications, such as live stream announcements and timed giveaways, is better scheduled for the evening for this audience.
More news for September 28, 2026
Grok Bot is a set of AI agents from xAI, Elon Musk's company that owns X. xAI introduced them last month: an agent performs tasks for the user, each bot has its own computer, it logs into the person's services and inboxes by itself, and comes back only for approval. A template is a ready-made bot workflow that other people can use. The pilot is invite only. A creator publishes templates on X and gets paid through X Money, the platform's payment service, when people use those templates. X decides whether to pay and how much in each payment cycle: it looks at how many people use a template and how regularly, and runs by the creator themselves don't count. X hasn't disclosed payout amounts. For people who know automation, this is a new way to earn on X, and the money will come from templates people keep coming back to, not from ones they try once.
Muse is Meta's personal AI agent, built into Instagram and Facebook, that performs tasks for the user. Instagram posted a short overview of how Muse helps creators make content and run their accounts. When a creator gives Muse access to their Instagram data, the agent starts offering growth advice: it suggests who to collaborate with, compares incoming partnership offers against industry averages, and puts together a creative plan for the account. It also produces summaries of the account and of its own actions, and sends regular reminders. The risk is that the agent can publish something that breaks Instagram's rules, and the creator's account gets penalized. That's why Muse has action approval: the agent does nothing without a human check. Instagram didn't name prices or timelines.
Along with the agent launch, Meta started a big PR campaign meant to convince people their data is safe with it. Muse agents need serious access: they search the web, make purchases, analyze banking data, pick insurance plans, and sort through health information. The outlet traced complaints against Meta back to 2007. In 2026, hackers tricked Meta's AI support bot and got access to more than 20,000 Instagram accounts. In total, lawsuits like these have cost the company more than $25 billion in fines. Meta promises to change its approach: a new system called Private Processing will handle data from its devices inside "confidential virtual machines." For now the system is a promise, while the agent is already live. So before connecting Muse to an ad or business account, cut its permissions to the minimum.
An API is the interface third party services use to connect to a platform. The new API lets outside video platforms launch and manage live streams on X. According to the company, it was rewritten from scratch. You can now schedule a stream, moderate the chat, and go live through various streaming providers. The feature set includes persistent RTMP streams (the standard protocol programs like OBS use to send video to a platform), creating, publishing, and monitoring broadcasts, scheduling, reading and replying to chat, and moderation with the ability to mute or remove a viewer. Before this, in July, X launched a live stream control center in Creator Studio, the creator dashboard inside X. Now you can run streams on X through your usual streaming service without going into the app itself, which makes it easier to broadcast to several platforms at once.
What creators, advertisers, and brands should do about it
- Creators and brands with a US teen audience on TikTok: check your posting times against the restrictions. The night hours from 0 to 6 a.m. and school hours may drop out of your reach, and other states may follow Alabama.
- Anyone connecting Muse to a creator or brand account: turn on action approval right away and don't give the agent access to payments or the ad account unless it actually needs it.
- Anyone with an invite to the X pilot who knows how to automate routine work: publish Grok Bot templates that people will use every week. Payouts depend on the number of users and how regularly they use it, and your own runs don't count.


