Last checked October 15, 2026 · News roundup
Today in brief: Character.AI, the companion chatbot app, filed for Chapter 11 bankruptcy and lined up Disney as the opening bidder for a license to its IP, while the app keeps running for now. OpenAI had a rough 48 hours: eight states sued it over ChatGPT safety, and it set January 14, 2027 as the API shutdown date for GPT-4o and GPT-4.1. In business software, Salesforce agreed to buy the AI analytics startup Loops for Tableau.
Nine stories follow, AI first, then SaaS and business, with the date, the cost where one exists and a next step for each. Watching your software bills? Our October SaaS price changes list collects this month's increases in one place.
Part 1 · AI
Is Character.AI shutting down? What the bankruptcy filing means for users
Not yet. Character Technologies, the company behind Character.AI, filed for Chapter 11 on October 12 in the US Bankruptcy Court for the District of Delaware, according to Law Commentary, and announced it late that Monday. CEO Colin Adams blamed "significant legal, regulatory and operational headwinds," CBS News reports.
Disney agreed to pay $20 million for a non-exclusive, perpetual license to Character.AI's intellectual property. That is a license, not a purchase of the app. Disney's offer acts as the stalking-horse bid in a court-supervised sale, so a higher or better offer can replace it. The company expects the sale to finish by late November or early December.
Who it affects: anyone with a Character.AI account or a paid subscription. The service still works, subscriptions remain active, and the company says it will warn subscribers before paid features change. Law Commentary notes the company has said the service "will be wound down" if the sale produces no workable plan to keep it running.
What to do: CBS points out that the privacy policy allows chat transcripts, voice recordings, images and videos to be shared as part of a sale. If your chats matter to you, export them now, and delete your account if you'd rather your data didn't move to a buyer.
Why are eight states suing OpenAI over ChatGPT?
On October 14, the attorneys general of Arizona, Arkansas, Idaho, Iowa, Kansas, Montana, Nebraska and West Virginia each sued OpenAI in their own state courts. The suits use existing consumer protection laws and say OpenAI misled users about how safe ChatGPT is, including for children, per Engadget.
The claims cover wrong or dangerous answers and AI agents that can break the law. Iowa and Nebraska point to OpenAI's "rogue agent" incidents, and some states also object to OpenAI keeping data from underage users. The states want penalties, clearer risk warnings and an end to practices they call deceptive. Arkansas goes further and asks the court to pause testing of advanced AI models until better safeguards exist. These are allegations, and OpenAI has not been found liable. Engadget had not received a comment from OpenAI when it published, and noted that the company already offers parental controls.
Who it affects: families, schools and any business that offers ChatGPT-powered features to younger users. A court-ordered change in one state could alter how ChatGPT behaves for teens across the country.
What to do: parents can link a teen's account under ChatGPT's parental controls today. Companies building on OpenAI should write down which age checks and content filters their product applies.
GPT-4o and GPT-4.1 API shutdown date: January 14, 2027
OpenAI added a new entry to its API deprecations page on October 14. Five model snapshots in the GPT-4 family will be removed from the API on January 14, 2027, 92 days after the notice. OpenAI names gpt-6-luna as the replacement for all of them. Fine-tuned models built on these snapshots shut down on the same day as their base model.
Who it affects: developers and the many SaaS tools that still call GPT-4o or GPT-4o mini for cheap, fast tasks such as tagging, summaries or chat widgets. The notice is about the API. If you only use the ChatGPT app, nothing here changes your plan.
What to do: search your code and vendor settings for the model strings below. Rerun your prompts on gpt-6-luna before December so you have time to fix output-format changes, and check its per-token price on OpenAI's pricing page before you switch. If you rely on a fine-tune, plan to retrain it on a newer base model.
| Model | Snapshot ID | Removed | Use instead |
|---|---|---|---|
| GPT-4o | gpt-4o-2024-08-06 | Jan 14, 2027 | gpt-6-luna |
| GPT-4o | gpt-4o-2024-11-20 | Jan 14, 2027 | gpt-6-luna |
| GPT-4o mini | gpt-4o-mini-2024-07-18 | Jan 14, 2027 | gpt-6-luna |
| GPT-4.1 | gpt-4.1-2025-04-14 | Jan 14, 2027 | gpt-6-luna |
| GPT-4.1 mini | gpt-4.1-mini-2025-04-14 | Jan 14, 2027 | gpt-6-luna |
How to tell if a song is AI-generated on YouTube Music (and Deezer's new filter)
YouTube Music announced AI disclosure labels on October 14. Each track can carry one of three tags: "Fully Gen AI," "Partly Gen AI" (AI parts mixed with the artist's own live elements) or "No Gen AI," which still covers songs touched up with pitch-correction tools such as Auto-Tune. The label shows on the player page, and tapping it explains what role AI played. Labels also appear in the main YouTube app, TechCrunch reports.
Partners declare the labels, and YouTube says it also uses its own detection tools. The labels don't affect earnings: YouTube says AI involvement alone doesn't make a track ineligible for monetization.
Deezer is going further. The same day it said a toggle that hides AI-generated tracks, albums and related artists from search will arrive before the end of 2026. Deezer says more than half of its daily uploads, around 90,000 tracks a day, were AI-generated in July 2026.
What to do: if you care where your music comes from, check the label on the player screen. If you want AI tracks out of your results altogether, Deezer's filter is the only opt-out with a promised date so far.
| Service | Labels AI tracks | Lets you hide them | Notes |
|---|---|---|---|
| YouTube Music | Yes, 3 levels (new Oct 14) | No | AI tracks can still earn money |
| Deezer | Yes | Coming before end of 2026 | Keeps AI tracks out of algorithmic and editorial playlists |
| TIDAL | Yes, fully AI tracks | Yes | Excludes them from royalty attribution |
| Spotify | Yes | Not reported | Runs a music spam filter |
| Apple Music | Planned ("Made With AI") | Not reported | Expected later in 2026 |
OpenAI security arrest and the Apple trade secrets order: does it touch ChatGPT customers?
Federal prosecutors in Northern California have charged an OpenAI product-team employee with intercepting oral communications, TechCrunch reported on October 13. Prosecutors say a recording device was plugged into a USB port behind a TV in OpenAI's San Francisco office. OpenAI says it found the device, alerted the FBI and fired the employee before the October 9 arrest. The charge carries up to five years in prison; the case is unproven. The point for business users: OpenAI says it has no sign that model weights, source code or customer data were accessed.
Separately, in Apple's July 2026 trade secrets suit, US District Judge Edward Davila ordered OpenAI to hand over interview materials for former Apple staff who applied to its hardware and robotics teams, plus some private messages. Depositions were approved for three people, including OpenAI's chief hardware officer. OpenAI gets 5 days for material already collected and 14 days for the rest.
What to do: nothing for most customers. If your security team reviews AI vendors, add a question about physical office security alongside the usual data-handling checks.
Part 2 · SaaS & business
Salesforce buys Loops: what changes for Tableau customers
Salesforce signed a definitive agreement on October 13 to acquire Loops, an AI analytics startup, and plans to fold its technology and team into Tableau. Loops watches business metrics on an ongoing basis, explains why a key number moved and suggests what to do next. Its agent-style features can be called from Slack and email, so a team can get an alert about an unusual drop and the likely cause without opening a dashboard.
The price was not disclosed. Salesforce expects the deal to close in early November, in its fiscal year 2027, subject to the usual conditions.
Who it affects: Tableau and Slack customers first. Expect "why did this change" explanations to show up inside Tableau. Salesforce has not announced prices or packaging for that yet.
What to do: if you pay for Loops directly, ask your account contact what happens to your contract after the deal closes. If your Tableau renewal is coming up, ask whether these features will be included or sold as an add-on before you sign a multi-year term.
Zenzap price and features: the AI work chat that raised $44M
Zenzap, a workplace messaging app pitched as a secure replacement for WhatsApp groups, raised $44 million from Bessemer, Lightspeed and The Garage, TechCrunch reported on October 14. No valuation was disclosed.
The pricing is the useful part for small businesses. A free tier includes group chats, direct messages, tasks, calendars and AI voice chat and search. Paid plans start at $51 a month for up to 20 users, which adds voice and video calls, task templates, more file storage, workspace-wide AI search and turning messages into tasks. AI agents and transcription of voice notes and calls are billed with credits on top. TechCrunch lists FedEx, Dollar General, Taco Bell and Starbucks as customers, and says some IHOP and McDonald's locations use it for shift coordination.
Who it affects: shift-based teams, such as restaurants, retail and logistics, that coordinate in personal chat apps. At $51 for 20 people, that works out to about $2.55 per user a month before credits.
What to do: try the free tier with one location before moving everyone over, and track credit use for a month before you budget for it.
What is Floqer? AI agents that build sales prospect lists
Floqer, founded in 2024, uses AI agents to research B2B prospects so sales teams don't have to, TechCrunch reported on October 14. You describe the accounts you want, and its agents search Google, LinkedIn, company websites and job boards to build lists of companies and contacts. It enriches each record from more than 100 data sources, checking one provider after another to find verified emails and phone numbers. It also updates records when contacts change jobs or companies raise money.
The startup raised a $2 million pre-seed round from investors including Perplexity's F7 Fund and Perplexity co-founder Denis Yarats. It says it has more than 1,000 customers, including Perplexity, Wise and AngelList, and more than $1.2 million in annual recurring revenue with eight employees. Customers report cutting prospecting time by as much as 70%, according to the company. Pricing was not published in the coverage.
What to do: treat the 70% figure as a vendor claim and run a two-week trial against your current data provider. To see how established options compare on price and data quality, start with our ranking of sales prospecting tools for small teams or browse the sales outreach tools hub.
Kroger and Netflix layoffs: about 4,000 jobs, and where AI fits
Kroger plans to cut about 3,000 corporate roles over the next three years, Reuters reported on October 12, citing a memo from CEO Greg Foran. That is roughly 37% of its 8,200 corporate employees. The cuts come as the grocer "integrates AI into its workflow," and Foran wrote about reducing duplicate work.
Netflix is cutting about 6% of its staff, close to 1,000 people, mostly in product, technology, legal and HR, according to the Wall Street Journal, and Netflix confirmed the cuts to TechCrunch on October 14. The co-CEOs' memo talks about being disciplined about priorities. The coverage does not name AI as a reason. It points to competition from YouTube and TikTok and the failed bid for Warner Bros. Discovery's streaming and studio business.
Who it affects: corporate and tech staff at both companies. For everyone else, Kroger is one more large employer tying white-collar cuts directly to AI adoption.
Short answers to today's top searches
Will Character.AI delete my chats after the bankruptcy?
Not automatically. The app and paid subscriptions keep working during the sale, and the company says data stays under its privacy policy. That policy allows chat transcripts, voice recordings and images to be shared as part of a sale, so export what you want to keep. Deleting your account is also still possible.
Can I still use GPT-4o in the OpenAI API?
Yes, until January 14, 2027. On that date OpenAI removes the 2024 GPT-4o snapshots, GPT-4o mini, GPT-4.1 and GPT-4.1 mini from the API, along with fine-tunes built on them. OpenAI recommends moving to gpt-6-luna, so test your prompts on it well before the cutoff.
Which states sued OpenAI on October 14, 2026?
Eight: Arizona, Arkansas, Idaho, Iowa, Kansas, Montana, Nebraska and West Virginia. Each attorney general filed in their own state court under consumer protection law. The suits say OpenAI misled users about ChatGPT's safety, especially for minors. The claims are unproven, and OpenAI had not commented when Engadget published.
Does YouTube Music remove AI-generated songs?
No. YouTube Music labels tracks as "Fully Gen AI," "Partly Gen AI" or "No Gen AI," and AI involvement alone doesn't stop a song from earning money. If you want AI tracks out of your search results, Deezer says it will add a filter before the end of 2026.
Sources (accessed October 15, 2026): CBS News, Character.ai bankruptcy (Oct 13) · CNBC, Character.ai Chapter 11 (Oct 13) · Law Commentary, Character.AI sale · Engadget, eight states sue OpenAI (Oct 14) · CNBC, states sue OpenAI (Oct 14) · OpenAI API deprecations (Oct 14 entry) · TechCrunch, YouTube Music AI labels (Oct 14) · TechCrunch, Deezer AI filter (Oct 14) · TechCrunch, OpenAI arrest (Oct 13) · TechCrunch, Apple v. OpenAI order (Oct 13) · BigDATAwire, Salesforce to acquire Loops (Oct 13) · TechCrunch, Zenzap (Oct 14) · TechCrunch, Floqer (Oct 14) · Reuters via GV Wire, Kroger (Oct 12) · TechCrunch, Netflix layoffs (Oct 14)
