While TikTok is often labeled as a major social media player, it has been progressively expanding its offerings. The video application has incorporated features like TikTok Shop, a local discovery map, advanced search functionalities, games, and much more. Most recently, the platform introduced hotel booking options and is in the process of obtaining a fintech license. 

It seems that TikTok is making strides to transform into a “super app,” a unified platform where users can engage in activities beyond merely watching and sharing videos, addressing a variety of user needs in a single location. 

The super app concept has gained significant traction in China, exemplified by apps like WeChat, which serves as a blend of Facebook, WhatsApp, Apple Pay, and an app store. Naturally, one might wonder if the super app model can thrive outside of China, but that does not deter TikTok from attempting it.

Rather than hopping between different applications, TikTok is striving to become the go-to app for users conducting a variety of digital tasks. Following its major development with TikTok Shop, the company — which notably transitioned to primarily U.S. ownership last January — has applied the same approach to its recent advances. 

Sports

Through the years, TikTok has consistently sought to establish itself as a destination for sports highlights and related content. In early June, the platform unveiled a dedicated hub for the FIFA World Cup, allowing users to access scores, match schedules, standings, trending videos, player clips, and more.

If users are already consuming videos on TikTok but wish to stay updated on scores, they can do so without exiting the app or resorting to a separate sports news app or a search engine.

The World Cup hub was enabled through TikTok’s sports-oriented product, “TikTok GamePlan,” which aims to enhance discovery and engagement for sports teams, leagues, and broadcasters on the platform. The social media leader also maintains partnerships with Major League Soccer (MLS) and Major League Baseball (MLB) for exclusive behind-the-scenes content.

Hotel and attraction booking

In May, TikTok rolled out TikTok GO, an initiative that allows users to discover and book hotels, attractions, and experiences directly within the app in the United States. TikTok GO showcases accommodations and activities via videos, search functions, and location pages. When users find something of interest, they can access details, check availability, and finalize a booking. 

Rather than redirecting users to external sites after they locate a destination or suggestion within a video, TikTok is positioning itself as a comprehensive platform where viral travel-related content can facilitate bookings and revenue generation.

While users have already turned to TikTok as a substitute for Google for searches, this new venture places TikTok in direct competition with Google’s primary services, such as Search and Google Maps, as it aims not only to be the app users rely on to explore locations but also the venue where they book their journeys. 

Payments

In March, it was reported by Reuters that TikTok had sought approval from Brazil’s central bank to function as a financial technology entity offering lending and payment services. 

The organization is pursuing two licenses. The first would permit it to provide prepaid accounts for users to save funds, receive payments, and execute transactions. The second license would enable it to operate as a direct credit provider, allowing it to lend its own resources or serve as a platform linking borrowers and lenders. 

This initiative signifies a notable shift for TikTok as it transcends its identity as merely a social media entity and ventures into the digital ecosystem. By integrating financial services into its app, TikTok aims to enhance user engagement, explore new revenue sources, and establish itself in competition with fintech startups and e-commerce platforms.

TikTok Shop

It’s well-known that TikTok’s major progress beyond social media was marked by the introduction of TikTok Shop. The platform began testing TikTok Shop in 2021 and fully launched it in the United States in 2023. Since then, the company has managed to effectively compete with Amazon, Shein, and other online shopping platforms.

According to eMarketer, TikTok Shop experienced a remarkable 407% growth in U.S. sales in 2024, followed by an additional 108% increase in 2025, reaching $15.82 billion. By last year, the company held an 18.2% share of total social commerce in the U.S., a figure projected to rise to 24.1% by 2027.

Moreover, TikTok further challenged digital marketplaces with the introduction of TikTok Shop gift cards late last year. The platform has also recently ventured into luxury retail, expanding its initial reputation for affordable goods.

Music

TikTok’s rise has significantly impacted the music landscape and how audiences discover new tracks, prompting the company to launch a streaming service named TikTok Music in 2023 to compete with giants like Spotify and Apple Music. However, it was discontinued within a year. 

The organization noted it would pivot towards promoting music listening and continuing partnerships with existing streaming services rather than competing directly with them. However, TikTok has not entirely retreated from its music initiatives, as it recently introduced a feature allowing Apple Music subscribers to play full tracks within the app after finding them in their “For You” feed.

Search and Maps

TikTok has introduced a comprehensive search feature that highlights maps, local hashtags, and reviews to assist users in discovering popular restaurants, travel destinations, shops, and local events. Furthermore, it has provided more thorough details about locations and eateries on dedicated pages, allowing users to rapidly access information such as operating hours, ratings, price bands, and more.

TikTok has already been encroaching upon Google’s search domain since its inception, quickly surfacing videos showcasing commentary and visuals of food and places. Nevertheless, users might have still resorted to Google Search for specific locations or review insights. In recent years, however, TikTok has progressively diminished that necessity by incorporating detailed place information directly within its platform.

Microdramas

While TikTok is praised for user-generated content, it has also started exploring microdramas by creating an in-app Minis section and a standalone app for short-form TV shows that can be enjoyed through a series of one-minute episodes. Although TikTok already vies for user attention with streaming services like Netflix, its foray into scripted programming intensifies its competition.

It’s noteworthy that TikTok first ventured into entertainment through live streaming and support for extended video lengths, a pivotal shift from its original focus on 15-second clips. 

Games

TikTok has also introduced a range of casual games within its app to encourage users to spend additional time on the platform and connect with others via direct messages. The inclusion of games signifies TikTok’s ongoing transition from merely a social media network to a comprehensive entertainment hub where users can not only browse videos but also engage in friendly game challenges.

The U.S. administration is preparing to exert significant control over the release of various AI models.

Just two weeks after the U.S. government halted the deployment of Anthropic’s Fable and Mythos models, OpenAI’s upcoming model appears to be facing similar delays. The Information reported on Thursday that GPT 5.6 will be available only in a limited preview, with the government approving the access “customer by customer” until a broader release is permitted.

If that preview only continues for a “couple of weeks,” as Altman allegedly suggested, it may not pose too much of an issue. However, Mythos has already been in preview for several months, and there’s no sign that a general release is imminent. Even a few weeks in review could greatly hinder the economic benefits of an expensive new system, especially as AI labs strive to enhance their financial performance. A slowdown in model development as a result could similarly impact ongoing data center expansions.

Should this situation worsen, the entire sector may face considerable peril.

Importantly, OpenAI and Anthropic now find themselves in an identical predicament, facing the same challenges and potential disasters if they fail. Discussions within the tech sector often highlight one party’s role in this situation, either blaming Anthropic for attempting to capture regulatory favors or accusing OpenAI of forming connections with Trump to undermine a competitor. This is understandable; many influential individuals in the industry have substantial financial stakes in one company or the other. 

However, what is unfolding now extends beyond these individual conflicts. The implications of instituting a chaotic government approval process for every pioneering model are evident, and there’s no solution that benefits one lab without also providing assistance to others. 

The immediate challenge is to establish a coherent release framework. While it’s reasonable for the government to evaluate models prior to their release (as is common for numerous consumer products) — but as GMU fellow (and soon-to-be OpenAI employee) Dean Ball articulated eloquently in a post this morning, it remains unclear what type of safety assurances could be implemented to meet regulatory expectations. The U.S. government lacks the specialized knowledge or resources for the level of testing that would be required. Additionally, it’s uncertain what risks regulators aim to safeguard against, as there has been no effort to clarify the government’s actual concerns.

It’s easy to perceive the government process as the entirety of the dilemma, but there are legitimate issues at play. Even if one doesn’t subscribe to the Mythos enthusiasm, there is clear evidence demonstrating how AI technologies are transforming cybersecurity. Similar dynamics are present in biorisk and alignment discussions. Limiting model releases cannot be the only solution — that will merely restrict public access — but there are serious issues that need attention.

The most effective approaches to address these issues, as outlined by Ball, will necessitate collaboration. It will involve placing trust in independent organizations to steer the process, even if their objectives do not fully align with your own. It will require rallying behind the least problematic regulatory alternatives instead of contesting every regulation with fervor. Most importantly, it will demand advocating for AI as an industry rather than perceiving safety and regulation as opportunities to gain a competitive edge.

For many individuals working in AI, this will be a challenging proposition. Regrettably, AI models have evolved to a stage where their capabilities carry tangible political ramifications. Addressing those ramifications will necessitate collective action. In the coming weeks, we will discover whether the industry is capable of such collaboration.

Peter Diamandis, the founder of Xprize Foundation, has aligned himself with a growing cadre of technology leaders who advocate for global surveillance, asserting, “[h]umans act more responsibly when they know they are being observed.”

This week, Diamandis voiced his views in a post on X and elaborated on his convictions in a Substack article, where he essentially characterized the concept as: Big Brother with positive intent.

“Radical transparency is on the horizon. A future where knowledge can be acquired anytime, anywhere. A future where no one remains hidden,” he expressed on Substack. “We are enveloping the Earth in a ‘Sensor Ecosystem’: a dynamic, multi-layered sensing apparatus that extends from home cameras to smartphones, autonomous vehicles, and humanoid robots on the ground, to drones and flying cars in the air, culminating with a network of satellites capturing images of every square meter on Earth daily.”

Diamandis’ remarks came about two years after Oracle founder Larry Ellison made similar statements.

“Citizens will behave better because everything happening will be continually recorded and reported,” Ellison forecasted during an Oracle event in 2024.

Diamandis seems to have been inspired to express these views following a podcast interview with Will Marshall, CEO of Planet, the largest operator of Earth-monitoring satellites.

“No one can elude detection anymore,” Marshall informed Diamandis during their discussion. “If you construct a school, we will observe the school. If you establish a data center, we will detect the data center. And the accountability will be visible to the entire world, regardless of circumstances.”

Diamandis, Ellison, and Marshall are correct that much of this technology is currently operational and expanding. It is becoming increasingly difficult for individuals to move through their daily lives without being captured on camera by home security systems like Ring, vehicles equipped with cameras like those produced by Tesla, or automated license plate recognition systems from Flock. Even if they manage, they are still monitored through their phones by advertising networks and data brokers.

However, Diamandis’ remarks are among the most candid regarding the intention to eliminate privacy.

“Your children will grow up in an environment devoid of ‘off the record’,” he advises any parents engaging with his post. “Educate them that the most effective privacy approach is to act with integrity, living in such a way that being observed incurs no cost. And advocate vigorously for a world where the observation is mutual.”

Diamandis appears to consider this an unavoidable reality, yet that perspective does not reflect how the average person is reacting to the surge of surveillance technology. Some municipalities have covered their Flock cameras with garbage bags following reports that the company’s data was being accessed by ICE, the FBI, and other law enforcement entities. Public opposition to Ring’s “Search Party” feature—designed to locate lost pets, a typically defensible initiative—led to the company terminating its collaboration with Flock.

Meanwhile, Meta has faced backlash regarding its camera glasses (created in collaboration with Ray-Ban), and is also embroiled in a lawsuit concerning privacy issues.

A considerable portion of Diamandis’ Substack post centers on providing guidance to entrepreneurs or executives on how to navigate a world lacking privacy. This advice primarily distills to: “be a good person.” Yet he struggles to address whether individuals will choose to act this way due to moral obligation or the possibility of surveillance. (He mentions that this is the question he’s “been pondering” since concluding the interview with Marshall.)

What Diamandis does not grapple with are the same fundamental queries that technology leaders frequently circumvent in discussions about surveillance and privacy. The interpretations of “good” or “honest” are often subjective—particularly in the context of powerful tech companies overseeing the surveillance framework.

Diamandis briefly posits that these firms promote transparency, asserting that “transparency is a tool, and tools lack ethics.” However, he overlooks the reality that tools often embody the biases of their creators. Who determines what behavior recorded by a security camera is deemed “good” or “honest”? This vital question remains unaddressed.

All he is willing to assert is that transparency “only cultivates trust when it flows both ways.” Achieving that equilibrium seems challenging, at best, within a landscape where the technology necessary for such “transparency” is controlled by a select few.

This evening is your final opportunity to secure up to $190 off your pass for TechCrunch Founder Summit 2026. The Early Bird special concludes tonight, at 11:59 p.m. PT, after which prices will rise.

Founders seldom grow in isolation. The quickest route to expansion lies in learning from those who have already succeeded, engaging with fellow entrepreneurs facing similar obstacles, and fostering connections with investors who can propel your next phase of development.

On November 4 in Boston, over 1,000 founders and investors will convene for a well-organized day filled with practical learning, open discussions, and significant networking aimed at assisting founders in making informed choices and accelerating their growth.

Don’t delay until prices go up.  Sign up today before 11:59 p.m. PT to enjoy savings of up to $190. Bringing your team along? Groups of four or more can benefit from savings of up to 30%.

An event for founders focused on growth

Every session, dialogue, and networking chance is crafted around the genuine challenges founders encounter while establishing, funding, and expanding their enterprises.

You’ll engage with:

• Founders navigating corresponding growth phases.

• Seasoned operators who have expanded teams, products, and revenue.

• Investors providing insights on what they are funding and what they seek.

Whether you’re getting ready to raise funds, fine-tuning your go-to-market plan, or outlining your next developmental milestone, the Founder Summit fosters dialogues that can significantly alter the path of your venture.

Practical insights for every phase

The Founder Summit centers on the choices that determine a startup’s trajectory. Through breakout sessions and roundtable discussions, you’ll acquire actionable advice on topics like:

• Securing Series A funding.

• Crafting pitch decks that attract investors.

• Understanding the right time to pursue Series C and beyond.

• Growing to $10 million ARR.

• Recognizing how and when to sell your business.

• Preparing for an IPO.

These discussions led by founders offer practical insights that you can implement right away.

Gain knowledge from founders and investors who’ve achieved success

Previous speakers have included:

Other speakers have come from Sequoia Capital, NFX, Glasswing Ventures, Wing Venture Capital, Construct Capital, Greylock, Precursor Ventures, and others.

The 2026 schedule is forthcoming, with additional founders, operators, and investors to be revealed.  Visit the event page for updates on the agenda.

Would you like to lead a discussion?  Put forth a topic for a roundtable or breakout session for the chance to be included on the agenda through audience voting at TechCrunch.

Secure up to $190 off before tonight at 11:59 p.m. PT

The cutoff is in less than 24 hours. Early Bird pricing concludes tonight at 11:59 p.m. PT. Join the founders, operators, and investors who are defining the next wave of startups. Acquire practical insights, establish valuable connections, and depart with strategies you can implement immediately.

Sign up tonight before 11:59 p.m. PT to enjoy savings of up to $190 on your pass and a group discount of up to 30%.

Considering exhibiting at TechCrunch Founder Summit 2026? Book your exhibit space to connect directly with founders, investors, and key startup decision-makers.

Wander through San Francisco, and you’ll quickly notice an empty self-driving car gliding through the streets, either waiting for a passenger or heading to a remote location for charging and servicing. These unoccupied trips — a term in the industry for distance traveled without a fare-paying passenger — represent one of the primary challenges facing robotaxi firms in achieving profitability.

Startup Aseon Labs, based in Redwood City, California, believes it has a solution: automated pods the size of parking spaces that can be deployed throughout urban areas to maintain, clean, and charge robotaxis. Co-founded by the creators of battery-swapping company Pushme, the firm refers to these innovations as robotic pit stops tailored for the robotaxi sector. The concept has drawn interest from investors.

Aseon Labs has secured $10 million in seed funding led by Crane Venture Partners, according to information obtained by TechCrunch. Contributions also came from Y Combinator, Garrett Camp’s venture firm Expa, Robin Hood Ventures, and Founders Capital, along with angel investors including serial entrepreneur and former Google executive Adrian Aoun, Mercury founder and CEO Immad Akhund, Zimride co-founder Rajat Suri, and founding team members from Anthropic, Nuro, Turo, and Revolut.

Currently in its initial phase, Aseon Labs intends to use the seed funding to develop five prototypes of these pods, expand its team of six robotics and engineering professionals to around a dozen, and acquire the necessary real estate to establish its network, as stated by CEO and co-founder George Kalligeros.

“To achieve economic parity with ride-hailing — which is our target for self-driving vehicles — and to cease heavily subsidizing costs, we need to increase utilization,” Kalligeros explained to TechCrunch. “We need the robotaxi actively operating throughout the entire daily demand cycle.”

Aseon’s approach suggests that a network of dispersed autonomous pods would significantly reduce unproductive miles, ultimately transforming robotaxi services into feasible business ventures.

Kalligeros and COO Dan Keene come from outside the realm of self-driving vehicles. However, they bring expertise from developing and scaling a hardware-and-real estate venture. Kalligeros was a mechanical design engineer at Bentley Motors and Tesla before co-founding Pushme in 2016 to create battery-swapping solutions for micromobility fleets. Pushme was establishing a battery-swap network in Europe when it was acquired by Tier Mobility in January 2020.

“The analogy I’ll make is that we were essentially tasked by SoftBank to deploy this across as many viable markets for Tier in an expedited timeframe,” Kalligeros noted. “The strategy became about how to distribute locations around the heart of the city sensibly while ensuring deployment was manageable as non-permanent infrastructure.”

Aseon Labs applies the same mindset to autonomous vehicles.

During their exploration of the sector, the duo visited AV depots where fleets of robotaxis undergo inspections, maintenance, cleaning, and charging. The high costs of real estate often lead businesses to position these depots away from city centers, where the majority of ride-hailing activity happens.

“Depot infrastructure is crucial for launching new cities for any AV operator,” he stated. “And what transpires in the depot at present — the essential framework for autonomy, really — isn’t fully developed.”

The founders focused on developing smaller, autonomously powered pods that can be placed throughout an urban landscape but also relocated as necessary. These units, equipped with cameras for vehicle inspections and robotic arms for retrieving lost items and interior cleaning, are categorized as temporary structures. This classification assists Aseon Labs in bypassing extensive permitting procedures and allows the company flexibility in relocating underperforming units.

The units are designed to operate on propane generators or other mobile power sources or can connect to existing power infrastructures through collaborations with EV charging firms. They are intended to function autonomously, though initial models will include staff, according to Kalligeros.

Aseon Labs is not attempting to address every potential scenario either. Instead, it utilizes computer vision and AI — particularly vision-language-action models prevalent in contemporary robotics — to identify issues that the pod should not attempt to resolve. For instance, if a camera identifies melted chocolate on a backseat, the robotic arm will refrain from intervening as trying to clean it could exacerbate the stain. Rather, the vehicle will be charged and sent directly to the company’s main depot for human intervention.

While Aseon Labs has yet to finalize agreements with any robotaxi companies, Kalligeros mentioned that there is considerable interest in the initiative. “Almost everyone is keen to give it a try,” he stated.

The introduction of OpenAI’s latest model, GPT 5.6, is said to differ from earlier editions. Rather than making it available to everyone, the company intends to distribute it exclusively among a limited number of trusted partners at the insistence of the Trump administration, according to The Information.

During a recent meeting, CEO Sam Altman reportedly informed employees that the government would be “granting access on a customer-by-customer basis” in a preliminary phase. Altman also mentioned that if this restricted launch is successful, OpenAI aims to proceed with a wider public release in “a few weeks.”

In essence, the Trump administration seems to be urging OpenAI to adopt a strategy that Anthropic is already implementing voluntarily: keeping its most advanced AI technologies confidential.

As per The Information, not only is OpenAI’s forthcoming model undergoing scrutiny by the administration, but its employees have also “collaborated closely” with governmental entities in preparation for the release. The federal offices requesting a restricted distribution include the Office of the National Cyber Director and the Office of Science and Technology Policy.

The Trump administration — which initially adopted a “hands-off” stance towards AI — has recently advocated for federal regulation of new AI models. Earlier this month, Trump enacted an executive order asking certain AI firms to voluntarily present new models to the government for analysis and review prior to public launch. 

Earlier this year, Anthropic generated significant debate when it declared that its new cutting-edge cyber model, Claude Mythos, would be accessible only to a select group of partners through a program dubbed Project Glasswing. Anthropic contended that its model was simply too potent and could, if misapplied, do more harm than good. Since then, discussions have emerged on whether Anthropic’s claims are a marketing tactic or a genuine effort to prevent misuse of a powerful model. The truth may lie somewhere in between.

Cybercriminals have long employed automated tools, but in the current era of generative AI, they possess an unprecedented arsenal. LLMs have demonstrated a capacity for crafting malware, with some being able to autonomously execute complete ransomware operations.

The primary apprehension regarding advanced cyber tools like Mythos is their purported ability to both detect and exploit software vulnerabilities at speeds unattainable by human analysts. Given that numerous software systems harbor concealed bugs that serve as gateways into corporate networks, this presents a clear and substantial challenge for any organization managing intricate software infrastructures. However, as these models remain inaccessible to the public, assessing the actual level of threat they pose is challenging.

YouTube is implementing a variety of updates to Shorts, including a fresh feature that enables users to reduce the length of their short-form videos.

The platform, owned by Google, revealed on Thursday that Shorts now includes an option that permits users to increase their playback speed. The aim of making an already brief experience even shorter is to help users “take in information more swiftly or locate your favorite segment quicker,” according to the platform.

In an apparent effort to foster a more positive online environment, YouTube has also eliminated the dislike button for Shorts. Now, instead of expressing disapproval for a video, users will need to use the “Not Interested” and “Don’t recommend this channel” features to deter specific types of content.

Likewise, rather than pressing a thumbs-up button if they enjoy a video, users will now be able to use a heart emoji.

Lastly, YouTube is introducing a new “Clear Screen mode,” which aims to temporarily remove “all icons and text from your playback view,” providing users with an unobstructed view of their content free from distracting overlays.

All of these modifications are aimed at creating “a more intuitive Shorts experience,” stated the company. It remains unclear when these updates will be officially rolled out. The company mentioned that the features will be implemented gradually, but specific dates were not provided.

TechCrunch contacted Google for further details.

YouTube debuted Shorts in 2020 and has successfully drawn a substantial audience since that time. As of June 2025, YouTube Shorts was reportedly averaging 200 billion views daily, as noted by CEO Neal Mohan during his keynote address in Cannes last year. (This noteworthy statistic can be contextualized with the fact that YouTube defines a “view” as the initial moment a video is played.)

A report earlier this year indicated that Shorts were increasingly being viewed on users’ TV screens — with consumption reaching as high as 2 billion hours of such content each month.

Correction June 26: This article has been updated to rectify the year Shorts was launched.