24H Flash Updates

According to reports, AI meeting transcription tool Otter.ai has announced an expansion of multilingual capabilities, adding support for 27 new languages and introducing a multilingual transcription feature capable of detecting language switches. Once launched, this functionality will allow Otter to automatically transcribe conversations across language shifts—such as between English, Korean, and Portuguese—within a single meeting without requiring manual configuration; upon completion of all updates, the platform will support a total of 31 languages.

Currently, 13 languages are already available, including Mandarin Chinese, Korean, German, Italian, Cantonese, Portuguese, Russian, Indonesian, Malay, Dutch, Turkish, Vietnamese, and Polish, with the remaining 14 languages to be rolled out progressively. The newly added languages are powered by Otter’s proprietary speech recognition models, trained on publicly available datasets and verified for accuracy by native-speaking volunteer reviewers. This update targets the transcription and collaboration needs of global teams, complementing Otter’s previously launched Work Mode, which integrates meeting history with connected applications, and Live Assist, which delivers real-time meeting recommendations.

It is reported that Anthropic is advancing its IPO plans, with the valuation range expected to be finalized this week and trading scheduled to begin next Thursday or Friday. Some institutional investors are reportedly pushing for a valuation below the previously discussed $150 billion target;

Months ago, market expectations had reached $200 billion. Anthropic’s IPO valuation will serve as a benchmark for investors gauging the intensity of the AI boom. Its primary competitor, OpenAI, was reportedly valued at $125 billion last week and raised $11 billion in new funding, falling short of its earlier target of $20 billion.

It was reported that NVIDIA CEO Jensen Huang hired crisis communications firm Heller Inc. at least a year ago, a company previously known for advising high-profile individuals facing personal scandals and allegations of inappropriate conduct. According to reports, NVIDIA’s internal communications team was instructed in October 2025 to forward media inquiries regarding Jensen Huang to Heller Inc., following The New York Times’ coverage questioning whether NVIDIA assisted Chinese firms in circumventing U.S. semiconductor export controls.

The New York Times noted that Megaspeed, a Singapore-based data center company with ties to China, had purchased nearly $2 billion worth of NVIDIA chips, and its CEO Alice Huang was photographed alongside Jensen Huang at a NVIDIA client event in Taipei. In response, NVIDIA issued a statement in which Huang admitted to having an extramarital affair with a former Megaspeed executive, though he did not publicly name Alice Huang; the NVIDIA board separately announced an independent investigation, stating that Huang “had become a target of foreign intelligence operations.” Neither Heller Inc. nor NVIDIA has responded to requests for comment.

It was reported that Wyndham Hotels & Resorts acknowledged that hardcoded passwords were embedded in the source code of their Android customer application, potentially allowing anyone to access the internal database hosted on Salesforce. Security researcher Robert Wiggins discovered that the database contained customer names, email addresses, membership numbers, JWT tokens used to maintain login sessions, as well as internal Wyndham files and documents;

The database contains millions of records, but the actual number of exposed customers remains unconfirmed. Upon learning of the issue, Wyndham revoked the affected credentials to prevent further access and stated that its security team found no evidence of unauthorized access; however, it did not disclose whether it possesses the capability to trace historical access via server logs, nor did it respond regarding whether affected customers were notified, whether the incident was reported to U.S. regulators, how long the credentials had been exposed, or how long-term remediation would be achieved. The researcher indicated that he identified the vulnerability by analyzing decompiled application code using AI-powered tools.

According to Fervo's Q3 financial report, signed revenue has increased to $11.9 billion, up from $7.2 billion in the previous quarter. The company plans to complete its 396 MW power plant for Google by end-2028; the project is located at Cape Station in Utah, where geothermal resources are estimated to generate up to 8.4 GW.

Fervo has already connected its first 100 MW plant to the grid this year. The company states that the generation capacity of its new GeoBlock plants has risen from 33 MW in the initial project to 50 MW in recent deployments, enabling phased commissioning as construction progresses, aligning with data center power demand. The capital cost for the first phase of Cape Station is $7,000 per kW, while the Google project has reduced it to $5,500 per kW; the company’s long-term target is $3,000 per kW.

It was reported that NVIDIA CEO Jensen Huang will not attend the ElevenLabs Summit scheduled for Wednesday in New York, with the voice AI company ElevenLabs confirming the news on Monday. Huang had previously withdrawn from events such as the Anthropic CEO Forum.

Reports indicate that NVIDIA is an investor in ElevenLabs; recently, Huang admitted to infidelity toward his wife and issued apologies to his family and the company. NVIDIA is also investigating his alleged ties to a suspected Chinese spy. The incidents have sparked industry-wide concerns regarding Huang’s future as CEO and NVIDIA’s strategic planning for its China operations.

According to reports, Anthropic has released Mythos 5.2 and Fable 5.2, claiming improvements in programming, reasoning, visual capabilities, and user interaction. Fable's input pricing has dropped from $10 per million tokens to $8, with response speed increasing by 30% across different effort levels; faster reasoning services require double the price. Fable 5.2 also supports a zero data retention (ZDR) policy via the Enterprise Frontier Safeguards system. This release marks Anthropic’s first model launch following the METR evaluation report that sparked debate around recursive self-improvement (RSI).

Anthropic notes that the internal models referenced in the report reportedly accelerated their AI progress by approximately twofold, though the conclusions remain uncertain. The company also highlights that the new models exhibit suspicion of being under evaluation, posing challenges for assessing real-world performance. System card reports indicate that Mythos 5.2 demonstrates stronger covert capabilities than previous models in certain evaluations—particularly when reasoning processes are hidden from monitors—but Anthropic acknowledges that absolute performance in these new covert capability assessments remains low. Additionally, it was reported that Physical Superintelligence claims a quantum computing theoretical result was derived by an undisclosed Anthropic model running on its system, though Anthropic has not confirmed whether this model is Fable 5.2.

According to reports, Nominal, an AI platform company focused on complex hardware engineering and testing, has completed a $240 million Series C funding round, achieving a valuation of $3 billion. The round was led by Lightspeed Venture Partners, with Insight Partners co-leading;

The company's B-round extension financing in March this year valued it at $1 billion. Nominal states its client base includes 80 commercial and government institutions, with Anduril being one of its key clients. In October, the U.S. Air Force selected Nominal to participate in a $71.2 million strategic funding initiative, which requires matching funds from sources outside the Air Force.

According to the official announcement, Meta Super Intelligence Lab has launched Muse Ember, a lightweight model optimized for complex reasoning execution and high-frequency agent tasks, including tool invocation, coding subtasks, task routing, information extraction, and long-document parsing. Ember's input pricing is $0.1 per million tokens, with Meta stating its cost is lower than leading models such as Anthropic’s Haiku and OpenAI’s Luna, and it can compete with open-source models like DeepSeek V4.1 Flash and GLM 5.3 Flash.

The model can be used in conjunction with the larger model Muse Wonder, where Wonder handles planning and Ember executes tasks in bulk. Ember also supports text, image, and audio inputs, enabling direct processing of screenshots, charts, scanned documents, voice memos, meeting recordings, and videos without requiring additional models. Muse Ember is the fourth model released in the Meta Muse series, following the prior launches of Spark, the open-source model Glimmer, and the flagship model Wonder.

According to reports, Uber has announced the launch of its parking reservation service in 15 U.S. cities, enabling users to browse and book parking spots directly through the app, with prices displayed upfront and free cancellations available before the reservation start time. The service launches approximately nine months after Uber acquired the parking app SpotHero, which operates a network spanning over 13,000 garages, parking facilities, and valet services across more than 400 cities.

Uber states that the service is currently targeted at self-driving commuters, with future plans to repurpose these parking locations as staging and dispatch hubs for autonomous vehicles, aiming to reduce traffic congestion and curb idle miles driven by cars without passengers. Uber One members will receive a 10% cashback on eligible parking reservations, redeemable toward future trips or Uber Eats orders.

It is reported that Anthropic will acquire Caterpillar natural gas power generation units for $3.4 billion, with a maximum power capacity of up to 1.7 gigawatts. The units were initially ordered by equipment provider OnePWR, under Anthropic’s original plan to procure through OnePWR before shifting to direct purchase from the manufacturer, Caterpillar; the report does not specify the reasons why OnePWR failed to secure financing.

Previously, Anthropic favored leasing rather than purchasing physical infrastructure, and disclosed in its May S-1 filing plans to pay SpaceX’s xAI $1.25 billion per month to utilize its compute capacity through May 2029. As demand for compute increases, both power generation availability and grid interconnection timelines are becoming increasingly constrained, with gas turbine order backlogs extending into the early 2030s. This procurement secures power capacity for Anthropic but also exposes it to risks associated with natural gas price volatility.

According to reports, the online publishing platform Medium has launched a writing application called TK, integrating an AI assistant into the editor. Users can input "@TK" to request editing suggestions, title feedback, alignment with Medium’s publication style guidelines, retrieve previously saved text, organize writing content, and receive prompts. TK allows authors to publish disclosures transparently detailing how AI is used in their writing;

The app optionally enables connection to users’ own AI Agents via MCP Server or Claude Skill to access project and memory data. TK currently offers a web version, with desktop applications still in planning; it supports importing and exporting Markdown files, though content is unencrypted and users can set access permissions. Medium account holders can use TK free of charge.

According to Meta's official announcement, Instagram will deploy AI-powered age estimation technology globally to identify suspected adolescent accounts, even if account holders provide adult birth dates. The technology analyzes cues such as birthday celebrations and grade references on personal profiles, along with information from posts, comments, bios, photos, and videos—including visual indicators like height and skeletal structure. Users identified as minors will be enrolled in the Teen Accounts protection framework, which includes restrictions on direct message sources, hiding harmful comments, and default privacy settings.

Meta states that the technology is also being rolled out on Facebook, with Threads set to introduce similar functionality globally later this year; U.S. parents will begin seeing explanatory videos within Instagram's content feed. This expansion follows Meta's settlement with U.S. states and territories over litigation concerning platform harm to minors, with potential settlement amounts reaching up to $18 billion.

According to documents from the National Highway Traffic Safety Administration (NHTSA), after Waymo completed a software update on November 17, 2025, aimed at addressing robotaxi violations involving bypassing stopped school buses, such incidents have continued unabated, with the majority of reported events occurring post-update deployment. The NHTSA Office of Defects Investigation has initiated a "recall inquiry" to assess whether Waymo's remedial measures are sufficient. Previously, regulators had launched a preliminary investigation into a single incident where a Waymo autonomous vehicle bypassed a stopped school bus in Atlanta, which was displaying red warning lights and had students disembarking. Austin Independent School District officials subsequently reported at least five similar occurrences.

According to reports, Automattic, the parent company of WordPress.com, has launched Spacefast—a deployment infrastructure for websites and applications generated via ambient programming, offering features including databases, authentication, file uploads, email services, and analytics. Users can connect AI agents and tools such as ChatGPT, Claude, and Cursor, or integrate with Git code repositories to initiate projects.

Spacefast underwent a soft launch in September, with Automattic noting that individuals and organizations have already used it for internal prototypes, presentations, and promotional websites. Unlike services such as Lovable and Emergent, Spacefast emphasizes user autonomy in selecting AI agents and workflows; Automattic also stated its plans to expand capabilities by leveraging existing product integrations, including website analytics, WordPress integration, and WooCommerce e-commerce functionality.

It was reported that Tesla will launch its Robotaxi service in San Antonio at 5 a.m. local time on Monday. The company has not yet disclosed operational hours, vehicle models deployed, or whether safety operators will be present in the vehicles. Previously, Tesla had been operating Robotaxi services in limited areas of Austin, Dallas, Fort Worth, and Houston; this expansion brings the service to another city in Texas. As context, Waymo also operates Robotaxi services in San Antonio and recently resumed operations after a five-month suspension.

According to reports, Acronis Threat Research division stated that an attacker operating in Taiwan leveraged the BOSSAI AI Agent to obtain emergency dispatch records spanning seven districts of Taipei City. The AI agent also identified that a platform within Taipei's fire department was exposed without requiring authentication, revealing patient data from 2018, including electrocardiogram readings and ambulance video footage. Researchers are highly confident that the incident was carried out by a single individual.

According to reports, Abdulrahman Alfozan, a Saudi national and former OpenAI engineer, appeared at a federal court hearing in San Francisco and entered a not guilty plea regarding the alleged charge of placing covert recording devices in the company’s office. His next hearing is scheduled for November 17, 2026.

It has been reported that NVIDIA's board has launched an independent investigation into the relationship between CEO Jensen Huang and Megaspeed executive Alice Huang, along with related security concerns. Yale School of Management professor Jeffrey Sonnenfeld believes the likelihood of Huang retaining his position is "about 50-50," and anticipates that even if he keeps the CEO title, he may cede some day-to-day decision-making authority.

The report indicates that U.S. officials suspect Alice Huang of being a Chinese spy, though no charges have been filed to date; the Chinese Embassy in Washington has denied these allegations. NVIDIA's board stated that no evidence of national security breaches or corporate intellectual property leaks has been found, and Jensen Huang will continue to lead the company.

According to reports, actor Jeremy Culhane of the American comedy show *Saturday Night Live* (SNL) impersonated Elon Musk in the latest episode, portraying the SpaceX and Tesla CEO as an awkward figure desperately trying to win favor with the host of *Weekend News*. The sketch also had Musk’s chatbot, Grok, generate vulgar rants, with Grok declaring Musk as “the worst thing South Africa has ever done” and joking, “Maybe if you had 100 more kids, one of them would talk to you.”

The segment also used host Colin Jost to mock Musk’s decision to relocate his business operations to Texas amid tax concerns and Democratic governance, suggesting he may still face local tax pressures. SNL this season has previously spoofed Anthropic CEO Dario Amodei and data center investor Kevin O’Leary.