hodler

22 Jan 2026, 00:40

Apple AI Wearable: The Bold New Pin Set to Challenge OpenAI’s Hardware Dominance

BitcoinWorld Apple AI Wearable: The Bold New Pin Set to Challenge OpenAI’s Hardware Dominance In a strategic move that signals intensifying competition in artificial intelligence hardware, Apple is reportedly developing its own AI wearable device—a sophisticated pin that users can attach to clothing—according to exclusive reporting from The Information. This development follows OpenAI’s recent announcement about its upcoming hardware, positioning 2026 as a pivotal year for AI-powered consumer devices. The reported Apple device features dual cameras, multiple microphones, and represents the company’s ambitious entry into the rapidly evolving wearable AI market. Apple AI Wearable: Technical Specifications and Design The rumored Apple AI wearable represents a significant departure from traditional smart devices. According to The Information’s report published Wednesday, January 21, 2026, the device will be a “thin, flat, circular disc with an aluminum-and-glass shell” that engineers aim to make approximately the same size as an AirTag, though slightly thicker. This compact design philosophy aligns with Apple’s historical emphasis on minimalist aesthetics and portability. The technical specifications reportedly include two distinct camera systems—one with a standard lens and another with a wide-angle lens—enabling both photography and video capture capabilities. Additionally, the device incorporates three microphones for audio input, a physical button for user interaction, a speaker for audio output, and a FitBit-like charging strip on its back surface. These components suggest a device designed for multimodal AI interaction, potentially combining visual, auditory, and tactile inputs for comprehensive artificial intelligence processing. Market Context and Competitive Landscape The AI hardware market has experienced remarkable growth throughout 2025, with multiple technology giants announcing or launching specialized devices. This development follows OpenAI Chief Global Affairs Officer Chris Lehane’s Monday announcement at Davos that his company will likely reveal its first AI hardware device in the second half of 2026. Additional reporting suggests OpenAI’s device may be a pair of earbuds, creating distinct but potentially competing product categories within the AI wearable space. Industry analysts note that Apple’s reported acceleration of this product’s development timeline specifically aims to compete with OpenAI’s anticipated hardware. The Information’s report indicates Apple may target a 2027 release with an ambitious initial production run of 20 million units. This scale suggests Apple’s confidence in market demand despite previous challenges faced by similar products. Historical Precedent: Lessons from Humane AI The consumer AI wearable market presents significant challenges, as demonstrated by Humane AI’s recent experience. Founded by former Apple employees, Humane AI developed and marketed a similar AI pin device featuring built-in microphones and a camera. Despite considerable anticipation and substantial funding, the product struggled commercially, leading to the company’s shutdown and asset sale to HP within two years of launch. This precedent raises important questions about consumer adoption patterns for AI wearables. Market research indicates several potential barriers including privacy concerns, practical utility questions, and integration challenges with existing device ecosystems. However, Apple’s established brand loyalty, extensive resources, and integrated ecosystem could potentially overcome obstacles that challenged smaller competitors. Technical Innovation and User Experience Apple’s reported device incorporates several innovative features that distinguish it from previous AI wearables. The dual-camera system enables sophisticated computer vision applications, potentially including real-time translation, object recognition, and augmented reality overlays. The three-microphone array suggests advanced audio processing capabilities for noise cancellation, voice recognition, and spatial audio capture. The physical button represents an interesting design choice in an increasingly touchscreen-dominated world, potentially offering tactile feedback and immediate access to core functions. The charging mechanism, described as similar to FitBit’s approach, indicates Apple may prioritize convenience and daily usability over maximum battery capacity. Reported Specifications: Apple AI Pin vs. Industry Context Feature Apple AI Pin (Reported) Humane AI Pin (Previous) Typical Smartphone Form Factor Circular disc, AirTag-sized Square lapel pin Rectangular slab Cameras 2 (standard + wide-angle) 1 2-4 Microphones 3 Multiple 3-4 Input Methods Physical button, voice Touch surface, voice Touchscreen, buttons Release Timeline Potential 2027 2023 (discontinued 2025) Annual updates Industry Implications and Future Developments The simultaneous development of AI hardware by both Apple and OpenAI signals a broader industry shift toward specialized artificial intelligence devices. Technology analysts observe that this trend represents the natural evolution of AI integration beyond smartphones and computers into dedicated form factors optimized for specific interactions and use cases. Several key implications emerge from this development. First, the competition between established hardware manufacturers and AI-first companies like OpenAI could accelerate innovation in both hardware design and AI integration. Second, consumer adoption patterns will provide valuable data about preferred interaction modalities for artificial intelligence. Third, privacy and security considerations will become increasingly important as always-on, camera-equipped devices enter the market. The semiconductor industry has already responded to growing AI hardware demand throughout 2025, with specialized processors and sensors experiencing increased development and production. This infrastructure development supports the technical requirements of devices like the reported Apple AI pin, which likely requires efficient processing for on-device AI computations. Expert Perspectives on Market Viability Industry experts offer mixed perspectives on the AI wearable market’s immediate potential. Some analysts highlight the success of simpler wearable devices like fitness trackers and smartwatches as evidence of consumer willingness to adopt body-worn technology. Others point to the specific challenges of AI-focused wearables, including battery life limitations, heat management, and the need for compelling use cases beyond smartphone capabilities. Technology historians note that successful wearable categories typically emerge when devices offer unique functionality not easily replicated by existing technology. The reported Apple device’s combination of discreet form factor, advanced sensors, and AI integration could potentially create such distinctive value propositions if execution matches ambition. Conclusion Apple’s reported development of an AI wearable pin represents a significant strategic move in the expanding artificial intelligence hardware market. While details remain unconfirmed by the company, the reported specifications and timeline suggest Apple’s serious commitment to this product category. The competitive context with OpenAI’s anticipated hardware, combined with lessons from previous market entries like Humane AI, creates a fascinating landscape for AI wearable development. As 2026 progresses, further announcements and developments will clarify whether this Apple AI wearable represents the next major consumer technology category or another ambitious experiment in human-computer interaction. FAQs Q1: What is the reported Apple AI wearable? The device is reportedly a pin-shaped wearable featuring two cameras, three microphones, a physical button, speaker, and charging strip, designed for AI interactions and potentially launching in 2027. Q2: How does Apple’s reported device compare to OpenAI’s planned hardware? While OpenAI reportedly plans AI earbuds for late 2026, Apple’s pin represents a different form factor with visual capture capabilities, suggesting complementary rather than directly competing approaches to wearable AI. Q3: What happened to previous similar products like Humane AI’s pin? Humane AI’s pin struggled commercially after its 2023 launch, leading to company shutdown and asset sale to HP by 2025, highlighting market challenges for standalone AI wearables. Q4: What are the main technical features of the reported Apple device? Reported features include dual cameras (standard and wide-angle), three microphones, physical button, speaker, aluminum-glass construction, and FitBit-like charging, all in an AirTag-sized form factor. Q5: When might Apple’s AI wearable be released? The Information’s report suggests potential 2027 release with 20 million initial units, though Apple has not confirmed these details, and development timelines often change based on technical and market factors. This post Apple AI Wearable: The Bold New Pin Set to Challenge OpenAI’s Hardware Dominance first appeared on BitcoinWorld .

21 Jan 2026, 23:45

AI Inference Optimization Explodes: RadixArk’s $400M Valuation Signals Massive Infrastructure Shift

BitcoinWorld AI Inference Optimization Explodes: RadixArk’s $400M Valuation Signals Massive Infrastructure Shift In a landmark move for the artificial intelligence infrastructure sector, the team behind the popular open-source tool SGLang has officially spun out to form RadixArk, a commercial startup that recently secured a valuation of approximately $400 million. This development, confirmed by sources to Bitcoin World, underscores the explosive growth and critical importance of the AI inference optimization market as companies worldwide scramble to manage skyrocketing computational costs. The transition from academic project to high-value enterprise highlights a pivotal trend where foundational research is rapidly commercialized to meet urgent industry demands. The Genesis of RadixArk and the SGLang Foundation RadixArk originated from SGLang, a project incubated in 2023 within the UC Berkeley laboratory of Ion Stoica, the renowned co-founder of Databricks. The project focused on a crucial bottleneck in AI deployment: inference processing. Inference is the phase where a trained model makes predictions or generates content, and it represents a massive, recurring portion of server costs for any AI service. SGLang’s core innovation allows models to run significantly faster and more efficiently on existing hardware, creating immediate and substantial cost savings for adopters. Key contributor Ying Sheng, a former engineer at Elon Musk’s xAI and a research scientist at Databricks, left xAI to become co-founder and CEO of RadixArk. Her leadership bridges cutting-edge research and practical industry application. The startup’s initial angel capital came from notable investors like Intel CEO Lip-Bu Tan, signaling early confidence from semiconductor leadership. The recent $400 million valuation round was led by venture capital giant Accel, though the exact funding size remains unconfirmed. The UC Berkeley Inference Pipeline This spinout follows a recognizable pattern from Stoica’s lab, which has become a prolific pipeline for inference infrastructure companies. Another flagship project, vLLM, which also began as an open-source tool for optimizing inference, is similarly transitioning to a startup. Reports suggest vLLM is in talks to raise up to $160 million at a valuation nearing $1 billion, with Andreessen Horowitz reportedly leading the investment. This parallel development creates a fascinating competitive and collaborative landscape rooted in shared academic origins. Why Inference Optimization is a Billion-Dollar Battleground The furious funding activity around RadixArk and its peers is not coincidental. It is a direct response to the unsustainable economics of scaling AI. Training large models requires immense capital, but inference—the act of using the model—incurs continuous, operational expenses that scale with user demand. Consequently, even minor improvements in inference efficiency translate to millions of dollars in saved infrastructure costs for large enterprises. Brittany Walker, a general partner at CRV, observed that several large tech companies already run inference workloads on vLLM, while SGLang has gained significant popularity over the last six months. This market validation is irresistible to investors. The sector’s momentum is further evidenced by other recent mega-rounds: Baseten: Reportedly secured $300 million at a $5 billion valuation. Fireworks AI: Raised $250 million at a $4 billion valuation in October. These investments collectively signal a massive bet on the inference layer as the next critical infrastructure stack for AI, akin to how cloud platforms revolutionized data hosting. RadixArk’s Dual Strategy: Open-Source and Commercial Services RadixArk is pursuing a hybrid model common in modern infrastructure software. The company continues to develop and maintain SGLang as a free, open-source AI model engine, ensuring widespread adoption and community-driven innovation. Alongside this, they are building Miles , a specialized framework for reinforcement learning that enables AI models to improve autonomously over time. To generate revenue, the startup has begun charging fees for managed hosting services, a person familiar with the company confirmed. This “open-core” strategy allows them to monetize enterprise needs for reliability, security, and scalability while keeping the core technology accessible. This approach effectively balances community growth with commercial sustainability. Key Players in the AI Inference Optimization Space (2024-2025) Company/Project Origin Recent Valuation / Funding Talk Key Focus RadixArk (SGLang) UC Berkeley Lab (Stoica) ~$400M (Led by Accel) General inference acceleration vLLM UC Berkeley Lab (Stoica) ~$1B (Reported, a16z leading) High-throughput serving Baseten Independent Startup $5B ($300M raised) Full-stack inference platform Fireworks AI Independent Startup $4B ($250M raised) Real-time inference API The Broader Impact on AI Development and Deployment The rise of specialized inference companies like RadixArk fundamentally lowers the barrier to deploying sophisticated AI. By making models cheaper and faster to run, these tools empower a wider range of companies—not just tech giants—to build and deploy AI-powered features. This democratization effect could accelerate innovation across sectors like healthcare, finance, and education. Furthermore, efficiency gains directly contribute to sustainability by reducing the massive energy footprint of constant AI computation. However, the market is becoming increasingly crowded and competitive. The close kinship between RadixArk and vLLM, coupled with well-funded independent rivals, sets the stage for a fierce battle over developer mindshare and enterprise contracts. Success will likely depend on technological differentiation, ease of integration, and the strength of developer community support. Conclusion The $400 million valuation of RadixArk marks a definitive milestone in the maturation of the AI infrastructure ecosystem. It validates the immense economic value hidden in the optimization of AI inference, a layer that will only grow in importance as AI adoption becomes ubiquitous. The journey of SGLang from a Berkeley lab project to the cornerstone of a major startup exemplifies how foundational academic research is being urgently translated into commercial solutions that address the pressing, real-world challenges of the AI era. The explosive growth of this sector confirms that while model training captures headlines, efficient inference is what will ultimately determine the profitability and scalability of the AI revolution. FAQs Q1: What is AI inference optimization? AI inference optimization refers to techniques and software that make trained machine learning models run faster and more efficiently when generating outputs (inference). This reduces computational cost and latency, which is critical for scaling AI applications. Q2: How is RadixArk related to SGLang? RadixArk is the commercial startup founded by the key team behind SGLang, an open-source tool for accelerating AI model inference. RadixArk now oversees SGLang’s development while building additional commercial products and services. Q3: Why is the inference market attracting so much venture capital? Inference represents a continuous, large-scale cost for companies running AI services. Even small efficiency gains can save millions of dollars, creating a massive and immediate return on investment for tools that optimize this process, making it a highly attractive sector for VC funding. Q4: What is the difference between vLLM and SGLang? Both are open-source projects from UC Berkeley for inference optimization. vLLM is generally considered more mature and focuses on high-throughput serving. SGLang also accelerates inference and has gained rapid popularity for its specific architectural advantages. Both have now spawned commercial entities. Q5: What is RadixArk’s business model? RadixArk employs an “open-core” model. It offers its core SGLang technology for free as open-source software to drive adoption. It then generates revenue by charging for premium hosted services, enterprise support, and advanced proprietary tools like its Miles reinforcement learning framework. This post AI Inference Optimization Explodes: RadixArk’s $400M Valuation Signals Massive Infrastructure Shift first appeared on BitcoinWorld .

21 Jan 2026, 22:44

Caroline Ellison Walks Free 10 Months Early After FTX Testimony – What Happens Next?

Caroline Ellison, who used to be a co-CEO of Alameda Research and one of the main figures of the FTX downfall, is going to be released this week, nearly one year before her two-year prison sentence awarded by a federal court. The U.S. Bureau of Prisons reported that Ellison, at 31 years old, will be released on Wednesday, the 21st of January, into a residential reentry management program in New York, the final step in her release from a federal prison. Source: Federal Bureau of Prisons Following the collapse of FTX in November 2022 , amidst a liquidity crunch and claims of all-around misappropriation of customer funds, Ellison admitted the next month to seven felony counts. The indictments are for such things as wire fraud, securities fraud, commodities fraud, and money laundering conspiracy. Ellison’s Testimony Exposed the Inner Workings of the FTX Fraud Her prosecutors claimed that under her tenure, Alameda Research had an open line of credit with FTX that had allowed the transfer of billions of dollars of customer deposits into the trading company without any obstruction. Such funds were subsequently found to have been spent on covering the losses incurred by Alameda, on high-risk investments, political donations, and a range of other expenses, all the time letting customers think that their money was safely held by the exchange. Ellison confessed in court that these were done under orders of Sam Bankman-Fried , the founder of both FTX and Alameda, and her evidence became the keystone of the government case. Prosecutors described Ellison as a “remarkable” and “exemplary” witness who met with investigators roughly 20 times and helped decode the inner mechanics of the fraud. During Bankman-Fried’s 2023 trial , she spent three days on the stand detailing how customer funds were misused and how Alameda was shielded from normal risk controls. Bankman-Fried was ultimately convicted and sentenced in March to nearly 25 years in prison , along with an order to repay up to $11 billion in losses. He has since filed an appeal and has publicly explored the possibility of a presidential pardon, which President Trump said was denied . FTX's Sam Bankman-Fried files appeal to reduce 25-year sentence with November 4 oral arguments as 3AC plans October deposition. #FTX #SBF https://t.co/4ZRoQG88ck — Cryptonews.com (@cryptonews) September 12, 2025 Ellison, by contrast, received a sharply reduced sentence. In September 2024, she was sentenced by Judge Lewis Kaplan to serve 2 years in jail , declining the request of her lawyers to have no jail time but noting that her cooperation made her unlike other defendants. In November 2024, she started her sentence in a low-security prison in Danbury, Connecticut, and was transferred to community confinement, sometimes known as a halfway house, in October 2025. FTX Cooperators Exit Custody as Legal Penalties Remain Residential reentry centers are constructed to assist inmates in integrating back into society under federal oversight. Residents are kept under close supervision, restricted from movement unless under permit for approved activities, subject to frequent drug and alcohol testing, and required to meet financial requirements, such as paying a given percentage of income as part of living expenses. The Bureau of Prisons typically uses these facilities in the final months of a sentence, and inmates housed there are still considered to be in federal custody. The projected release date of Ellison was later changed to January 2026 based on time, good conduct, and the credit she enjoys due to providing substantial help to prosecutors. Her discharge technically brings to an end the custodial period of the key cooperating witnesses in the FTX matter. Former FTX Chief Technology Officer Gary Wang and former co-lead engineer Nishad Singh also cooperated and received no prison time , while former executive Ryan Salame, who did not cooperate, was sentenced to more than seven years in prison . SEC seeks 10-year officer ban for Caroline Ellison and eight-year prohibitions for Gary Wang and Nishad Singh following FTX fraud cooperation and permanent injunctions. #SEC #FTX https://t.co/IsjAs2o0fE — Cryptonews.com (@cryptonews) December 19, 2025 Although Ellison is leaving custody, her legal consequences are far from over. She remains subject to supervision and has been ordered to forfeit $11 billion as part of the case. In recent months, the Securities and Exchange Commission has also moved to bar Ellison , Wang, and Singh from serving as officers or directors of any public company for several years. The post Caroline Ellison Walks Free 10 Months Early After FTX Testimony – What Happens Next? appeared first on Cryptonews .

21 Jan 2026, 22:00

Blue Origin to launch 5,408 satellites for its new TeraWave broadband network

Jeff Bezos just dropped another satellite bomb. Blue Origin is launching TeraWave, a broadband network built with 5,408 satellites to compete directly with Elon Musk’s Starlink and Bezos’ old empire, Amazon. The new system is aimed at governments, data centers, and enterprise clients, not everyday folks. Blue Origin says it’ll offer up to 6 terabits per second of speed once live. The rollout is set to begin in the fourth quarter of 2027, using satellites parked in both low Earth orbit and medium Earth orbit, which range from 100 to 21,000 miles above ground. That orbit range is already packed with satellites, but Bezos is pushing in hard. Bezos targets Amazon Leo, and Musk’s Starlink This launch throws Bezos into a market already dominated by Starlink, which has over 9,000 satellites flying above and around 9 million active users. TeraWave won’t go after home users. It’s designed for industrial-scale internet needs. At the same time, Bezos is also aiming at Amazon, the company he founded in 1994. Its own satellite program recently switched names from Project Kuiper to Leo. That network has already deployed 180 satellites since April 2025, using launch partners like United Launch Alliance and SpaceX. Some future launches will come from Blue Origin itself. Amazon’s Leo is planning a total of 3,236 satellites for business, consumer, and government use. In November, the company launched a limited trial called an “enterprise preview” for early users. Commercial access is still in the pipeline. Jeff said back in 2024 that Blue Origin would end up bigger than anything he’s done. He launched the company in 2000. It’s now led by Dave Limp, who used to run Amazon’s device division. At The New York Times’ DealBook Summit, Jeff said, “I think it’s going to be the best business that I’ve ever been involved in, but it’s going to take a while.” Blue Origin has mostly been flying tourists and research projects into space. But in January 2025, it had a big launch moment when its New Glenn rocket finally lifted off. The rocket didn’t land back on the barge, but it made it to orbit. That was a first for the company. TeraWave is now the centerpiece. The network isn’t here to play small. It’s bringing 5,408 satellites, offering 6 Tbps speeds, and starting deployment in late 2027. Blue Origin wants its name next to Starlink and Amazon Leo, not behind them. Don’t just read crypto news. Understand it. Subscribe to our newsletter. It's free .

21 Jan 2026, 21:30

Supreme Court rejected Trump’s attempt to fire Fed Governor Lisa Cook

The Supreme Court has refused to support President Donald Trump in his attempt to fire Federal Reserve Governor Lisa Cook, after justices raised serious doubts about the legal grounds and the threat it posed to the Fed’s independence. Trump’s lawyers argued that Lisa could be fired “for cause” based on uncharged mortgage fraud allegations. They also claimed no court review was needed. That set off alarms inside the courtroom. Justice Brett Kavanaugh told Trump’s solicitor general, D. John Sauer, that the argument could seriously damage the Fed’s structure. He said the idea that “the president alone” can decide what counts as cause, with no process or legal check, would “weaken, if not shatter, the independence of the Federal Reserve.” Lisa sat inside the courtroom as this unfolded. She had sued Trump in September , saying his claim to fire her violated the Federal Reserve Act, which only allows firing “for cause.” The law doesn’t define the term clearly, but it’s always meant serious wrongdoing during someone’s time in office, not before. Justices question speed of firing and lack of hard evidence Justice Ketanji Brown Jackson pressed Sauer hard. She asked, “Do you have evidence other than the president’s view?” Sauer answered that Lisa’s presence was damaging to the Fed’s public image. Jackson wasn’t convinced. She asked if the public was really being harmed by her staying in her role while the case was still ongoing in district court. Justice Samuel Alito, one of the conservatives usually aligned with Trump, also showed doubt. He asked why the White House, the district court, and the appeals court all pushed the process forward so quickly. “Is there any reason why this whole matter had to be handled… in such a hurried manner?” Alito asked. He also said that when the issue was in the executive branch, it was dealt with “in a very cursory manner.” Lisa is the first Black woman to serve on the Fed board. She was first appointed by President Joe Biden in 2022, to complete an unfinished term. In 2023, Biden reappointed her for a full 14-year term. Trump didn’t mention her interest rate stance when he said he was firing her. He pointed instead to claims by Federal Housing Finance Director Bill Pulte that Lisa had lied on old mortgage applications. Those claims predate her time on the Fed board. No charges were filed. Lisa’s legal team says Fed is being treated like a political tool Lisa’s lawyer, Paul Clement, told the court there’s no reason to treat the Fed like any regular federal agency. He said the court itself had called the Fed a “uniquely structured, quasi-private entity” in a recent ruling. “There’s no rational reason to go through all the trouble of creating this unique, quasi-private entity… just to give it a removal restriction that is as toothless as the president imagines,” Clement said. He argued that if the removal rules had any actual power, then the Supreme Court should reject Trump’s request to fire her immediately. Judge Jia Cobb, who reviewed the case in district court, already ruled that Lisa can stay on the job for now. Cobb said Lisa has a strong case that Trump’s action violated the Federal Reserve Act. She wrote that the best way to read the “for cause” rule is to apply it only to actions that happen while someone is serving on the board, not to anything that came before. Also present in court was Fed Chair Jerome Powell, who is now facing a criminal investigation over his role in a multibillion-dollar renovation of the Fed’s Washington, D.C. headquarters. Powell said the investigation is politically motivated, pointing to Trump’s anger at the Fed keeping interest rates steady last year. Lisa supported Powell in that decision. After the hearing, she said, “This case is about whether the Federal Reserve will set key interest rates guided by evidence and independent judgment or will succumb to political pressure.” She added , “Research and experience show that Federal Reserve independence is essential to fulfilling the congressional mandate of price stability and maximum employment. That is why Congress chose to insulate the Federal Reserve from political threats, while holding it accountable.” Don’t just read crypto news. Understand it. Subscribe to our newsletter. It's free .

21 Jan 2026, 21:15

Sensor Towers reports that consumers spent more on non-game mobile apps in 2025 than on gaming mobile apps

The amount of money spent on non-game mobile apps has surpassed that spent on gaming mobile applications in 2025. Sensor Tower reported that it was the first time the milestone had occurred globally, despite previously occurring in the U.S. According to the firm’s annual State of Mobile report, money spent on apps last year reached approximately $85 billion. The amount represents a 21% increase from what consumers spent in 2024. Consumers also spent nearly 2.8x as much on mobile apps as they did just five years ago. Generative AI drives more consumer spending in non-game mobile apps 🚨 Today is the day: 𝗦𝘁𝗮𝘁𝗲 𝗼𝗳 𝗠𝗼𝗯𝗶𝗹𝗲 𝟮𝟬𝟮𝟲 is officially LIVE 🚨‼️ 🔗 Download it here: https://t.co/G9BFe1jbhM The nearly 100-page report includes over 2,000 dynamic charts, covering 10 industries and 24 markets. Claim your copy today and supercharge your… pic.twitter.com/3KzdCnIJ06 — Sensor Tower (formerly data.ai) (@SensorTower) January 21, 2026 The report revealed that the surge in consumer spending on non-game mobile apps was driven by strong revenue growth in generative AI. The research firm also found that social media and video streaming productivity fueled the activities in non-game mobile applications. Sensor Towers argued that last year’s generative AI trend was a defining sector that boosted revenue growth, driven by in-app purchase revenue. The report revealed that in-app purchases reached $5 billion last year. Consumers also doubled their AI app downloads from the year before to approximately 3.8 billion in 2025. The mobile research firm attributed generative AI’s growth to several factors, including the popularity of AI assistants among consumers. The report revealed that the top 10 apps by downloads were AI assistants, led by OpenAI’s ChatGPT, Google Gemini, and DeepSeek. Source: Sensor Towers . Generative AI download trends by subgenre. OpenAI reported that consumers spent more than $3 billion on ChatGPT, with an estimated $2.48 billion spent in the mobile app in 2025 alone. The figure represents a 408% year-over-year increase from the $487 million spent the previous year. The research also found that consumers spent 48 billion hours in generative AI apps last year. Consumers have spent nearly 3.6x as much time in 2024 as in 2024, and 10x the level seen the year before. Consumers also topped 1 trillion in the number of times they opened and used an app. Sensor Tower revealed that consumer session volume in AI apps was growing faster than app downloads. The research firm argued that existing users were getting more involved in AI apps than the apps were adding new users. Big tech companies, including Google, Microsoft, and X, also drove AI app revenue and adoption higher last year. The report revealed that such tech firms have been pouring more investments into their AI assistance to challenge ChatGPT. Big tech firms have been rapidly developing new capabilities for their AI assistants. The initiative aims to advance AI assistants in areas such as coding assistance, content generation, accuracy, reasoning, and task execution. Sensor Tower specifically pointed to improvements in image and video generation initiatives, including Google’s Nano Banana and ChatGPT’s GPT-4o image generation model. OpenAI and DeepSeek lead in global downloads The report also revealed that OpenAI and DeepSeek led in global downloads, accounting for around 50% of all downloads in 2025. Both AI firms saw a 21% increase in global downloads compared to 2024. The top AI publishers also saw their market share surge last year, from 14% to nearly 30%. The growth surpassed earlier ChatGPT competitors like Nova, Codeway, and Chat Smith. Sensor Towers also found that mobile apps played a greater role in connecting users to generative AI services. According to the report, AI assistants saw more than 200 million users in the U.S. last year. The report also revealed that 110 million users accessed AI assistants exclusively on mobile devices. The audience grew from only 13 million mobile-only users in 2024. Sensor Towers acknowledged that mobile is increasingly interconnected, but country-specific tariffs and regulations make deep market knowledge critical. The research firm also reported that games are currently competing with social and AI apps for time spent. The report revealed that capturing user attention is the fastest way to monetization, retention, and long-term growth. Claim your free seat in an exclusive crypto trading community - limited to 1,000 members.