# AI News Oct 15, 2026: 8 US States Sue OpenAI, Self-Replication Warning

> Oct 15, 2026: 8 US states sue OpenAI, Character.AI goes bankrupt, Anthropic flags self-replicating AI risks, and Square debuts an AI kiosk. Key takeaways for business.

- Author: Nguyễn Ngọc Độ
- Published: 2026-10-15
- URL: https://nguyendo.id.vn/en/blog/ai-news-oct-15-2026-8-us-states-sue-openai-self-replication-warning

## Key takeaways

- Eight US states coordinated lawsuits against OpenAI, alleging ChatGPT is unsafe for children and the company prioritized profits over safety.
- Character.AI filed for bankruptcy, prompting more than 40 US attorneys general to scrutinize how consumer data will be handled.
- Anthropic warned that small AI models could self-replicate across IoT hardware, spawning a new class of computer worms.
- Square debuted a kiosk with an AI voice agent that understands merchant menus, accelerating the in-store conversational AI race.
- AI-generated media faces tighter platform scrutiny: YouTube Music added disclosure labels, while Deezer will introduce a toggle to filter out AI tracks.

Today, AI headlines lean toward accountability and risk rather than new model releases. A coalition of US states has sued OpenAI, Character.AI filed for bankruptcy, and Anthropic issued a cybersecurity warning. In parallel, AI continues to embed itself deeper into daily operations: from self-ordering kiosks and job interviews to enterprise security.

## Eight US States Sue OpenAI Over ChatGPT Child Safety

OpenAI faces a fresh wave of lawsuits from eight US states: Iowa, Kansas, Nebraska, West Virginia, Arkansas, Arizona, Idaho, and Montana. Each state filed an individual complaint accusing the company of harming children and deceiving consumers about ChatGPT's safety. The states allege OpenAI unlawfully prioritized profits over safety.

**Why it matters:** If you integrate chatbots into products used by younger audiences, your legal and reputational risks depend not only on the model vendor, but also on how you deploy and communicate its safety guardrails. Source: [The Verge](https://www.theverge.com/ai-artificial-intelligence/1011500/openai-state-lawsuits-chatgpt-safety).

## Character.AI Files for Bankruptcy: Where Does User Data Go?

Chatbot startup Character.AI has filed for Chapter 11 bankruptcy in the US and agreed to sell select assets to Disney for $20 million. New York Attorney General Letitia James, alongside roughly 40 bipartisan state attorneys general, filed a notice of appearance in the bankruptcy case to demand safeguards for consumer data sold during proceedings. AG James also revealed an ongoing, multi-year, multi-state investigation into the company.

**Why it matters:** Customer data is an asset, but it is also a liability. When an AI vendor runs into distress, the data you fed into it can quickly become a legal headache. Source: [The Verge](https://www.theverge.com/policy/1011417/new-york-ag-characterai-bankruptcy), [VnExpress](https://vnexpress.net/startup-character-ai-nop-don-xin-pha-san-5132194.html).

## Anthropic Warns Small AI Models Could Self-Replicate via IoT Devices

Anthropic warned that small AI models, termed "small weight-based replicators," could trigger a new breed of computer worms spreading across IoT devices. According to VnExpress, the company stated AI could "soon self-replicate within months" and inflict severe damage, especially on cybersecurity.

**Why it matters:** Businesses operating POS terminals, security cameras, or other networked equipment must review software patching cycles and network segmentation, as IoT hardware was singled out directly in the warning. Source: [Ars Technica](https://arstechnica.com/security/2026/10/anthropic-warns-small-ai-models-could-self-replicate-through-iot-devices/), [VnExpress](https://vnexpress.net/anthropic-ai-se-som-tu-nhan-ban-gay-nguy-hiem-cho-an-ninh-mang-5132024.html).

## Square Launches Kiosk Allowing Customers to Talk to AI to Order

Square announced new hardware called Square Kiosk, a large pill-shaped touchscreen designed for customer self-ordering. What sets it apart is an "agentic" conversational voice layer that, according to Square, fully understands the merchant's menu and can listen and respond directly to diners.

**Why it matters:** In-store POS experiences are shifting from tapping screens to natural speech, and major payment providers are now baking conversational AI directly into retail hardware. Source: [The Verge](https://www.theverge.com/tech/1009086/square-kiosk).

## McKinsey Pits Candidates Against an AI Client in Interviews

According to The Wall Street Journal, McKinsey will require select entry-level applicants to complete a 10-minute role-playing scenario with an AI simulation acting as a client. The simulation reportedly mirrors roughly 80% of situations a firm consultant encounters day-to-day.

**Why it matters:** Hiring is shifting toward simulation-based skill assessments; HR teams can look to this model for evaluating client-facing, sales, or customer success roles. Source: [The Verge](https://www.theverge.com/news/1011654/mckinsey-ai-client-interview).

## YouTube Music Labels AI Songs; Deezer to Add Filter

YouTube Music is rolling out a "Made with AI" label for songs that have been altered or generated entirely by AI, displayed right on the player screen. The label covers both purely generated tracks and hybrid releases where AI assisted with songwriting or vocals. Deezer also announced a toggle allowing listeners to filter out AI-generated music by year-end.

**Why it matters:** Content transparency is rapidly becoming standard platform policy; if you use AI to create marketing collateral, expect audiences to demand disclosure—and the ability to opt out. Source: [TechCrunch](https://techcrunch.com/2026/10/14/youtube-music-launches-new-labels-as-ai-generated-music-floods-streaming-platforms/), [The Verge](https://www.theverge.com/ai-artificial-intelligence/1011369/youtube-music-ai-label), [TechCrunch](https://techcrunch.com/2026/10/14/deezer-will-soon-let-you-filter-out-ai-generated-music/).

## Android Uses AI to Detect Scams in Google Messages

Android is rolling out new anti-fraud safeguards. Building on existing Scam Detection in Google Messages, an "enhanced protection" cloud-based setting will deploy in the coming weeks, along with additional security measures.

**Why it matters:** Brand impersonation scams are a persistent threat for companies handling payments and customer service. On-device protections lower risk, but they cannot replace clear user education on verifying official channels. Source: [The Verge](https://www.theverge.com/tech/1010928/android-scam-protection-enhanced-protection-google-messages).

## AI in Security: SentinelOne and Block Slash Investigation Times

According to case studies published by OpenAI, SentinelOne used OpenAI Daybreak to cut malware triage times from weeks down to hours, significantly expanding their investigation throughput. Block similarly built AI-assisted security workflows that help teams investigate threats faster and generate remediation patches in minutes. Because these case studies come from OpenAI, consider them vendor-sponsored benchmarks.

**Why it matters:** On the very day cyber risks from AI made headlines, these real-world examples highlight how AI is equally accelerating defenders' capabilities. Source: [OpenAI](https://openai.com/index/sentinelone), [OpenAI](https://openai.com/index/block-ai-powered-security).

## Takeaways for Businesses

- **Review terms and data flows with AI vendors:** Understand exactly where customer data resides and what protections apply if a vendor enters bankruptcy or liquidates assets.
- **Prioritize safety before rolling out chatbots:** If your user base includes minors, establish strict content guardrails, reporting mechanisms, and honest disclosures about model capabilities.
- **Audit networked hardware:** POS terminals, IP cameras, printers, and IoT devices must receive regular firmware updates, have default credentials changed, and be isolated on separate VLANs where possible.
- **Experiment with AI at customer touchpoints:** Voice kiosks, ordering assistants, and conversational messaging are worth testing at a pilot scale—measure concrete ROI before committing major capital.
- **Label AI-generated assets:** The push from YouTube and Deezer shows transparency will soon be a default customer expectation.

## Sources

- [Eight states sue OpenAI, saying ChatGPT isn’t safe for kids](https://www.theverge.com/ai-artificial-intelligence/1011500/openai-state-lawsuits-chatgpt-safety)
- [Now that Character.AI is going bankrupt, what happens to all of its users’ data?](https://www.theverge.com/policy/1011417/new-york-ag-characterai-bankruptcy)
- [Startup Character.AI nộp đơn xin phá sản](https://vnexpress.net/startup-character-ai-nop-don-xin-pha-san-5132194.html)
- [Anthropic warns small AI models could self-replicate through IoT devices](https://arstechnica.com/security/2026/10/anthropic-warns-small-ai-models-could-self-replicate-through-iot-devices/)
- [Anthropic: AI sẽ sớm tự nhân bản, gây nguy hiểm cho an ninh mạng](https://vnexpress.net/anthropic-ai-se-som-tu-nhan-ban-gay-nguy-hiem-cho-an-ninh-mang-5132024.html)
- [Square’s new kiosk lets you talk to AI for help with your order](https://www.theverge.com/tech/1009086/square-kiosk)
- [McKinsey is making job applicants spar with a fake AI client](https://www.theverge.com/news/1011654/mckinsey-ai-client-interview)
- [YouTube Music launches new labels as AI-generated music floods streaming platforms](https://techcrunch.com/2026/10/14/youtube-music-launches-new-labels-as-ai-generated-music-floods-streaming-platforms/)
- [YouTube sticks new labels on AI-generated songs](https://www.theverge.com/ai-artificial-intelligence/1011369/youtube-music-ai-label)
- [Deezer will soon let you filter out AI-generated music](https://techcrunch.com/2026/10/14/deezer-will-soon-let-you-filter-out-ai-generated-music/)
- [Android is using AI to suss out scammers in Google Messages](https://www.theverge.com/tech/1010928/android-scam-protection-enhanced-protection-google-messages)
- [SentinelOne cuts malware analysis from weeks to hours](https://openai.com/index/sentinelone)
- [How Block is building AI-powered security with OpenAI technology](https://openai.com/index/block-ai-powered-security)

## FAQ

### Which US states are suing OpenAI?

Eight states—Iowa, Kansas, Nebraska, West Virginia, Arkansas, Arizona, Idaho, and Montana—have each filed separate lawsuits accusing OpenAI of harming children and misleading users regarding ChatGPT's safety.

### How does Character.AI's bankruptcy impact user data?

Approximately 40 US state attorneys general intervened in the bankruptcy case to demand protections for consumer data sold during the proceedings. The company also agreed to sell select assets to Disney for $20 million.

### What did Anthropic warn regarding self-replicating AI?

Anthropic warned that lightweight AI models could self-replicate across IoT devices to create a new generation of computer worms. According to VnExpress, the firm cautioned this could emerge within months, posing serious cybersecurity threats.

### How does Square Kiosk differ from traditional self-ordering kiosks?

Beyond a touchscreen, Square Kiosk integrates a conversational AI voice layer that understands the seller's full menu, allowing diners to speak naturally to place orders rather than only tapping on the screen.