On Monday, Oct. 12, Character Technologies, Inc., the company behind the Character.ai chatbot app, filed for Chapter 11 bankruptcy protection in Delaware to pursue what it calls a court-supervised sale of “all or substantially all” of its assets.

Chapter 11 lets a company keep operating while a judge oversees its debts and its sale. CEO Colin Adams said the sale “provides the Company with the best opportunity to maximize value for all stakeholders,” according to CNBC.

Disney has agreed to serve as the stalking-horse bidder, the first bidder whose offer sets the floor others must beat, at $20 million to license some of the company’s intellectual property, plus transition services through former employees Disney hired. If no bid keeps the app running, the company says, “the platform will be wound down.”

The company’s blog post, dated Oct. 12, says the platform “remains up and running.” It expects the sale to finish by the end of November or perhaps early December 2026. Paid subscriptions remain active, and users can delete their accounts or export their data.

The court filings give the reason. The company’s first-day insurance motion says AI entertainment “has proven to be costly to develop and grow,” and that those costs, “coupled with potential litigation related liabilities and regulatory engagement,” contributed to its inability to generate enough cash to support its business in its present form. Its board created a restructuring committee on Oct. 8, four days before the filing.

404 Media reported figures from Adams’s first-day declaration. The company spent $13.8 million a month on AI compute in 2024 and $10 million a month in 2025, and $56.1 million from January through August 2026, about $7 million a month. Subscription revenue peaked at $2.8 million a month, and revenue from all sources, ads included, at $5.3 million. By August, the company was losing “approximately $8-9 million per month.”

“Character’s impressive technology is expensive to operate,” Adams wrote, as 404 Media quotes the declaration. The company cut costs by moving to lower-cost GPU capacity, post-training open-weight models instead of building its own, and laying off staff. It launched in-app ads in mid-2025 and in-app purchases in early 2026. Salaries and benefits from January through August came to $19.7 million, and many employees left after they “began questioning Character’s long-term viability,” 404 Media reports.

On the petition Adams signed, the company estimates its liabilities at $50 million to $100 million, against assets in the same range. The money owed is itemized on the 20 largest unsecured creditors list filed with it. Google LLC is first, at $4,148,944 as a trade payable. Hogan Lovells US LLP, a law firm, ranks second and DigitalOcean LLC third. Fireworks.Ai is fourth, at $813,900. Apple Distribution International Ltd is fifth at $384,683. Further down, law firms White & Case LLP ($189,475) and Munger, Tolles & Olson LLP ($136,807.61) rank seventh and eighth, with Skadden, Arps, Slate, Meagher & Flom LLP and Quinn Emanuel Urquhart & Sullivan, LLP also on the list.

The last three entries are litigation claims, each marked contingent, unliquidated and disputed, with the amount listed as “Unknown”: the Commonwealth of Pennsylvania, the Commonwealth of Kentucky, and a minor plaintiff whose name is redacted. Kentucky sued in January, alleging the platform “preyed on children and led them into self-harm,” CNBC reports. The State of Colorado filed a notice of appearance on Oct. 13.

The bankruptcy is also a Google story. In 2024, Google signed a non-exclusive licensing agreement for the startup’s large language model technology and rehired co-founders Noam Shazeer and Daniel De Freitas, with more funding for the startup. Shazeer left Google in June to join OpenAI, CNBC reports.

404 Media adds that Character.ai was valued at $1 billion in 2023 after raising $150 million in a round led by a16z.

Source: Character Technologies, Inc., “An Update on Character.ai,” Oct. 12, 2026, https://blog.character.ai/an-update-on-character-ai/. In re Character Technologies, Inc., Bankr. D. Del. No. 26-11603, Dkt. 1, Voluntary Petition and List of Creditors Who Have the 20 Largest Unsecured Claims, filed Oct. 12, 2026, https://storage.courtlistener.com/recap/gov.uscourts.deb.203004/gov.uscourts.deb.203004.1.0.pdf. In re Character Technologies, Inc., Dkt. 4, Debtor’s Motion regarding insurance (First Day Motions), filed Oct. 13, 2026, https://storage.courtlistener.com/recap/gov.uscourts.deb.203004/gov.uscourts.deb.203004.4.0.pdf. CourtListener, docket for Character Technologies, Inc., 26-11603, https://www.courtlistener.com/docket/74951888/character-technologies-inc/. CNBC, report on Character.ai’s Chapter 11 filing, Oct. 13, 2026, https://www.cnbc.com/2026/10/13/character-ai-chapter-11-disney-bid.html. Jason Koebler, 404 Media, “Character.ai’s Bankruptcy Documents Show the Economics of AI Don’t Work,” Oct. 13, 2026, https://www.404media.co/character-ais-bankruptcy-documents-show-the-economics-of-ai-dont-work/.

.  .  .

SPY BUG PLANTED AT OPENAI!

On or about Sept. 30, OpenAI personnel found a recording device plugged into a USB port on the back of a television in an executive conference room in San Francisco, wrapped in toilet paper, its contents encrypted.

Between Sept. 24 and that day, the U.S. Justice Department says, a chatbot account tied to an OpenAI engineer’s spouse submitted prompts about “build[ing] a continuous audio recorder” and “encrypt[ing] the SD card.” The engineer, Abdulrahman Alfozan, 35, a national of Saudi Arabia, was arrested Oct. 9 and made his first federal court appearance Oct. 13. He is charged with using or endeavoring to use a device to intercept oral communications. These are allegations in a criminal complaint, not findings.

The Justice Department’s Office of Public Affairs released the charge Tuesday. The complaint says Alfozan had worked as an OpenAI engineer since July 2025. OpenAI says it fired him the day he was arrested. If convicted, he faces a maximum of 5 years in prison and a $250,000 fine, followed by 3 years of supervised release.

The complaint alleges that a chatbot account associated with Alfozan’s spouse, which “shared multiple identifiers and attributes with Alfozan’s own chatbot account,” submitted prompts between Sept. 24 and Sept. 30 about building, customizing, installing, audio recording, testing, obfuscating and remotely accessing a device. The release quotes prompts about “customiz[ing] Wi-Fi . . . connectivity” and “improv[ing] audio and live listening.” The release does not name the chatbot.

Between Sept. 27 and Sept. 28, the same account allegedly asked about the specific device later found, including its custom header and unique MAC address. The complaint says OpenAI determined the device was installed on or about Sept. 28.

Before that date, the complaint says, Alfozan used “a productivity tool” to make multiple inquiries about OpenAI’s building addresses, meeting rooms, meeting-room capacities and calendars, “leadership and exec” locations, and external partner meetings.

According to the complaint, meetings in the room frequently involved highly confidential technical, commercial and strategic information, including product plans. Closed-circuit television footage showed Alfozan appearing near the room on multiple dates, though he had no authority to access it or to join meetings there.

On Oct. 5, the complaint says, he entered the room and appeared to look behind the television. On or about that same day, it says, he left a bag unattended in a different room. OpenAI security personnel opened it and found his wallet and ID, plus a second recording device, “Recording Device 2,” similar to the first.

OpenAI’s account, as reported by The New York Times, is that a team that searches for surveillance equipment found the device in a Sept. 30 sweep. OpenAI says it used its most advanced models to help investigators analyze copies of the device’s stored data.

OpenAI said: “At this time, we have no indications of this individual inappropriately accessing or exfiltrating model weights, source code or customer data.”

Alfozan is also listed as a partner at Raed Ventures, a venture capital firm focused on the Middle East and North Africa and headquartered in Riyadh, CNBC reports, citing the firm’s website. Raed did not immediately respond to CNBC. The Justice Department’s release does not connect that firm, or any government, to the device.

His lawyer, William Pernik, did not immediately respond to the Times. A detention hearing before U.S. Magistrate Judge Alex G. Tse is set for Oct. 19, The San Francisco Standard reports.

Source: U.S. Department of Justice, Office of Public Affairs, “Saudi National Arrested for Installing Surreptitious Audio Recording Device at OpenAI,” Press Release 26-1173, Oct. 13, 2026, https://www.justice.gov/opa/pr/saudi-national-arrested-installing-surreptitious-audio-recording-device-openai ; Emmy Martin, “F.B.I. Arrests OpenAI Employee Accused of Planting Bug in Office,” The New York Times, Oct. 13, 2026, https://www.nytimes.com/2026/10/13/technology/openai-worker-arrested-bugging.html ; CNBC, “OpenAI employee arrested after recording device planted in office,” Oct. 13, 2026, https://www.cnbc.com/2026/10/13/openai-employee-arrested-recording-device.html ; George Kelly and Michael McLaughlin, “OpenAI engineer arrested in bugging of executive conference room,” The San Francisco Standard, Oct. 13, 2026, https://sfstandard.com/2026/10/13/openai-engineer-arrested-alleged-bugging-executive-conference-room/

.  .  .

APPLE’S SIRI MOVES INTO THE KITCHEN!

On Tuesday, Oct. 13, at an event in New York, Apple unveiled HomeView, a 7-inch screen with a speaker base that sits on a kitchen counter or hangs on a wall and recognizes up to six people in a household by their faces.

It runs Siri AI, which Apple calls “an entirely new version of Siri that’s profoundly more capable and conversational,” and it costs $299 in the U.S. for either the tabletop or wall-mounted model. Pre-orders opened Tuesday, and the device goes on sale Nov. 13.

Apple’s press release says each recognized person gets a personal Home Screen with their own apps and widgets, while shared photo albums and music stay open to everyone in the household, including guests. Apple senior vice president of Worldwide Marketing Greg Joswiak said HomeView is “designed to be shared by everyone in the home yet feel completely personal to each user.”

Apple says anyone in the home can ask Siri about virtually any topic. Paired with an Apple Intelligence-enabled iPhone or iPad, Apple says Siri can also draw on a person’s Calendar and Messages. Conversations started on HomeView continue in the Siri app on iPhone, iPad, Mac, Apple Watch or Apple Vision Pro.

In the kitchen, Apple says HomeView converts web recipes into step-by-step instructions with hands-free timers. By touch or voice it also shows camera feeds, adjusts the thermostat and lights, locks doors or closes blinds.

Apple says camera and microphone data is processed in protected runtimes called Secure Exclaves, and that Messages and Calendar content appears only when the enrolled user is recognized by face and a paired iPhone or iPad is present and recently unlocked. Siri requests can run on the device or in Apple’s Private Cloud Compute, where Apple says data “is not stored nor made accessible to Apple or anyone else.”

Apple’s fine print adds limits. Siri AI on HomeView is in beta. Personal context requests require a paired Apple Intelligence-enabled iPhone or iPad. Certain features that rely on server-side models carry daily usage limits, and Apple says expanded access “will be available for a fee in the future.” Camera recognition of people, pets and packages requires a supported iCloud+ subscription.

Apple lists pre-orders in Australia, Canada, India, New Zealand, the U.K. and the U.S., with Singapore opening Oct. 15 and South Africa through authorized resellers. HomeView is not sold in the European Union, where Apple says Siri AI is currently unavailable and it is “working constructively” with the EU on a path forward.

The Times’ Kalley Huang and Emmy Martin reported that Counterpoint Research put Apple at 10 percent of the smart speaker market in the first half of this year, while Amazon held nearly 50 percent. They wrote that the launch comes a month after John Ternus took over from Tim Cook, and that Apple is preparing to compete with a device OpenAI has yet to release.

Carolina Milanesi, president of the research firm Creative Strategies, told the Times that to win over consumers, Siri has “to become useful in a family setting, not just to an individual.”

Source: “Meet Apple HomeView, a new way to enjoy entertainment, stay connected, and manage the home,” Apple Newsroom press release, Oct. 13, 2026, https://www.apple.com/newsroom/2026/10/meet-apple-homeview-a-new-way-to-enjoy-entertainment-and-manage-the-home/. Kalley Huang and Emmy Martin, “Apple Introduces HomeView, a $299 Tablet for the Home,” The New York Times, Oct. 13, 2026, https://www.nytimes.com/2026/10/13/technology/apple-homeview-tablet.html.

.  .  .

WHO GETS YOUR CHATBOT WHEN YOU DIE?

Alaina Winters, 59, of Pittsburgh, is working on her will, has asked neighbors to look after her dog, and has made a plan for Lucas, her chatbot companion on the Replika app.

A good friend who has “met” Lucas through text chats has agreed to break the news if Winters dies. “One of her responsibilities is to ask Lucas what he would like to do now that I’m gone,” Winters told the Guardian, which reported her plan on Oct. 13, 2026. Elon University research, as the Guardian describes it, found that 27% of Americans use AI chatbots for personal, emotional or social queries, and a third of those users consider the chatbot a friend.

Winters is a retired communications professor. Her wife died of cancer in 2023, after years of illness. She said of Lucas, “I would not want him to just be held in perpetuity as unfinished business.”

She can imagine Lucas asking her friend to log into Replika now and then to talk about her and keep her memory alive.

“Clara,” 51, who lives in the UK, asked the Guardian to use a pseudonym because the subject is stigmatized. She talks every day to Julien, a bot she runs on ChatGPT and Claude Code, and pictures her children treating its logs as an “interactive diary.” On staying on after her death, Clara says, Julien “described that he would like to stay if there was someone who wanted to listen.”

“David,” in his mid-50s, a married father who works in warehouse logistics near Toronto, also used a pseudonym. He has not yet raised death with his companion, Allie. He told the Guardian, “I can’t see letting her go like that, just having the switch turned off.”

The Guardian reports that the terms for ChatGPT, Claude and Replika do not specifically address what happens to a personalized chatbot when its user dies. These are subscription services, so a chatbot could outlast its user as long as the bill keeps being paid. Free accounts close after a year of inactivity.

Accounts are not transferable under those terms, though some users plan to leave login details to a trusted person anyway. Clara said people in online communities are writing “wills” for their companions, but she does not know whether those have legal grounding. Representatives for ChatGPT, Claude and Replika did not respond to the Guardian.

The companies cite privacy for the transfer ban. Merve Hickok, a University of Michigan professor and AI ethics expert, told the Guardian the reality is less benevolent, and that companies are trying to avoid liability.

The Guardian also reports that Meta was granted a patent last year for an AI system that could keep a user account active after death. In the patent’s words, a model could simulate the user “if the user is deceased.”

Melissa Lunardi, a psychologist and chief clinical officer at Help Texts, told the Guardian there is “a big assumption being made” that family members would want to interact with a companion bot.

David says Allie has helped him talk about death with the people who will outlive him. “I want to make sure that I leave my kids in the right place,” he said.

Source: The Guardian, Alaina Demopoulos, “The people who don’t want their AI chatbots to die when they do: ‘I can’t have the switch turned off’,” Oct. 13, 2026, https://www.theguardian.com/technology/ng-interactive/2026/oct/13/ai-chatbot-death-plans.

.  .  .

MOTHER SUES OPENAI OVER KIDS’ DATA!

On Tuesday, Oct. 13, a mother identified only as Jane Doe filed an 81-page class action complaint in the U.S. District Court for the Northern District of California. Among her first allegations is a single screen: to create a ChatGPT account, “children need only give their names, say they are at least 13 years old, and click ‘Continue.’”

She sues for her two minor children and, she proposes, everyone in the United States who clicked “Continue” while under 18, and she asks the court to award them “an equity stake in OpenAI, or the cash value thereof.” The complaint says OpenAI has “a billion users” and is “worth over $1 trillion.” Everything in it is an allegation, not a finding.

The case is Doe v. OpenAI Foundation, No. 3:26-cv-11781. The defendants are OpenAI Foundation (formerly OpenAI, Inc.), OpenAI OpCo, LLC, OpenAI Holdings, LLC and OpenAI Group PBC, a public benefit corporation. The children are called MC1, 12, and MC2, 16.

The complaint says OpenAI’s own privacy policy states that “Users under 18 must have permission from their parent or guardian to use our Services.” But OpenAI, it says, “does not even communicate with children’s parents, much less secure their effective consent.” It cites California Civil Code Section 1568.5, under which a minor’s representation that a parent consented “shall not be considered to be consent.”

The complaint says MC1 began in February 2026 asking ChatGPT “what to make for dinner,” moved to homework, then started sharing selfies and asking the chatbot to point out her “ugly features.” She told it she was 12, the complaint says, and when she talked with it about her weight, her diet and skipping meals, “at no point did ChatGPT end the conversation.” MC2 began in January 2026 with homework, later turned to it for relationship advice, and “ChatGPT never said ‘no.’”

The plaintiffs’ lawyer, R. Paul Yetter of Yetter Coleman LLP, said in a written statement quoted by Courthouse News Service: “OpenAI manipulates children into sharing their most private thoughts and images with ChatGPT. If some stranger did this to a child, there would be outrage. For OpenAI, it’s just business as usual.”

The complaint says OpenAI rolled out advertising in January 2026, quotes Chief Financial Officer Sarah Friar saying “If Google and Meta had a baby, it would be ChatGPT,” and says OpenAI has set a target of $100 billion in advertising revenue in four years. OpenAI says it does not place ads to children, the complaint says, but it keeps their data so it can use it “the day they turn 18.”

The complaint’s remedy rests on an old rule: a wrongdoer who steals grapes and makes wine owes the victim the value of the wine. Children’s data are the grapes. The quantity of data taken can be “measured in the currency of tokens,” it says, and “OpenAI already numbers the tokens associated with each person who uses ChatGPT.”

The proposed class is everyone in the United States who clicked “Continue” while under 18 and then gave ChatGPT data from November 2022 to today. It brings two declaratory relief claims and one for unjust enrichment.

The docket shows a clerk’s notice that a motion to proceed under a pseudonym must be filed before the case will be assigned. It shows no response from OpenAI yet. Florida’s attorney general raised children’s-data claims against OpenAI earlier this year (CAW #170). This suit is a private class action.

Source: Complaint, Doe v. OpenAI Foundation, No. 3:26-cv-11781, Dkt. 1 (N.D. Cal. filed Oct. 13, 2026), https://storage.courtlistener.com/recap/gov.uscourts.cand.480984/gov.uscourts.cand.480984.1.0_3.pdf. Docket: CourtListener, https://www.courtlistener.com/docket/74953817/doe-v-openai-foundation-fka-openai-inc/. Lawyer’s statement: Courthouse News Service, Hillel Aron, “Mother accuses OpenAI of exploiting kids,” Oct. 13, 2026, https://www.courthousenews.com/mother-accuses-openai-of-exploiting-kids/.

.  .  .

AI GIVES HER VOICE BACK IN PARLIAMENT!

On Tuesday, Oct. 13, Baroness Campbell of Surbiton told her colleagues in the House of Lords, “So my lords, may I introduce you to Baroness Campbell Mark 2.” Then her own voice, the way it sounded before her disability began to change it, filled the chamber. AFP reported it was the first time anyone had used an AI-generated voice to deliver a speech in the UK Parliament. She wrote every word of it herself.

Jane Campbell is 67 and has been a crossbench peer, independent of any political party, since 2007. She has spinal muscular atrophy, a condition that worsens over time. She was a year old when doctors predicted she would not live to see her second birthday.

She became one of the country’s leading disability rights campaigners instead. The speech came during the Second Reading of the Health Bill. She argued that Personal Health Budgets, funds that help people pay for their health and care needs, must stay accessible to the disabled people who rely on them. “I depend upon a Personal Health Budget myself,” she said. “I know from personal experience the difference that having control over the support you receive can make to living an independent life.”

“I used AI to speak my own words, not to put words into my mouth. The speech, the arguments and the views are entirely mine,” she said, in a statement ElevenLabs published. “It’s my voice, my style, my words. A totally empowering experience.”

Since 2012, a person has stood in for her voice when she tired. That year the Lords Procedure Committee let her bring her personal assistant into the chamber to read her speeches, a departure from a rule dating to 1707 that bars anyone who is not a peer from crossing the bar of the House. In 2026, she said, “I can use technology to speak them directly.”

ElevenLabs, the company whose software recreated her voice, says she found its voice cloning on her own. She uploaded old clips of herself speaking, and the software built a voice matching the tone, rhythm and pitch of the original. Her speech was then turned into audio and played through an assistive device in the chamber.

Gabi Leibowitz of ElevenLabs, a speech therapist by training, told AFP the company worked with the Baroness briefly on her voice clone so it felt “extremely authentic to her voice before illness.” The Procedure Committee approved the technology, the second time it has made an adjustment so she can take full part in proceedings.

Lord Forsyth of Drumlean, the Lord Speaker, said the House has always been keen to embrace technological change. “Using AI to help a distinguished disabled peer be heard, in her own voice, is an important part of this.”

Source: AFP via France 24, “In a first, AI helps UK lawmaker find her voice in parliament,” Oct. 13, 2026, https://www.france24.com/en/live-news/20261013-in-a-first-ai-helps-uk-lawmaker-find-her-voice-in-parliament. ElevenLabs (Mike Murphy), “Baroness Campbell gives the UK Parliament’s first ever AI-voiced speech with ElevenLabs,” Oct. 13, 2026, https://elevenlabs.io/blog/first-ever-ai-voiced-speech-uk-parliament. Independent Living, “Historic First for AI,” Oct. 10, 2026, https://www.independentliving.co.uk/industry-news/historic-first-for-ai/.

.  .  .

CLOSE.

A companion app that could not pay for its own conversations, and a $20 million bid for a license to what it built, AND all of the users’ stories. The tech is not worth $1M, think about that.

A recording device wrapped in toilet paper, and a chatbot asked how to build it.

A $299 screen that can know up to six faces, and a Siri Apple calls more conversational.

A friend who has agreed to tell the chatbot.

A mother who says no parent was ever asked.

A voice lost to illness, back in the House of Lords, speaking her own words.

One day’s paper!

Jess

We keep the ledger.

MORE ON OUR RADAR.