Pi Network has launched three products aimed at extending its services beyond its own ecosystem. The releases were announced as part of the group’s Pi2Day update.

The additions are SoloHost, a framework for locally run AI applications and distributed computing; Pi Sign-in, which lets third-party apps and websites use Pi accounts for login; and PiVerify, an identity verification service for external businesses and developers.

Each product is designed to connect Pi’s existing network with outside users and companies. The move expands Pi’s offering from a blockchain-based community platform into infrastructure for computing, authentication, and identity checks.

Local computing

SoloHost is an open framework on Pi Desktop that lets third-party developers build and list applications for local AI use cases. The beta version is now available.

The framework lets users run self-hosted applications on their own computers without setting up servers or Docker environments. Those applications can then be accessed from mobile devices through the SoloHost app in Pi Browser, while data remains on the user’s machine.

The framework includes Hermes, an open-source local AI application that processes tasks directly on the user’s device. The aim is to reduce reliance on cloud infrastructure and keep data local.

SoloHost is also being positioned as a route into distributed computing. Pi plans to introduce a production use case that would allow about 100 Node operators to contribute spare computing resources to complete AI tasks.

That proposal builds on a network of more than 420,000 user-operated Node computers, according to the company. Participating Node operators could be paid by third-party clients in Pi, creating another use for those machines beyond maintaining blockchain infrastructure.

Identity access

Pi Sign-in takes Pi’s account system beyond Pi Browser to third-party websites and desktop applications. The service allows users to sign in with their Pi account rather than create separate login credentials for each service.

For developers, the product is intended to provide access to Pi’s existing user base and verified identity system. It could simplify onboarding and give external platforms a way to authenticate users through Pi.

The sign-in product also links to SoloHost. A user could install an application locally on a desktop machine and then use Pi Sign-in to authenticate another device for remote access to that locally run software.

That means desktop software running through SoloHost can identify the right user and restrict access accordingly. The setup can also support interactions with other parts of Pi’s ecosystem, including wallets.

Verification service

The third release, PiVerify, opens Pi’s identity verification system to outside organisations. The service is aimed at businesses and platforms that need to establish whether users are genuine individuals.

Pi’s verification system has already been used to verify more than 18 million users across more than 200 countries and regions, according to the company. It combines AI and human review in its know-your-customer process.

PiVerify goes beyond document and liveness checks, the company said. It also includes sanction and anti-money laundering screening support, human validator workflows, cross-network comparisons, and support for different regulatory formats.

The service is being pitched to sectors including fintech, Web3, data, and AI. External clients could integrate it into onboarding or payment workflows to reduce fake accounts, duplicate registrations, and Sybil attacks, in which one participant creates multiple identities.

Wider strategy

The releases point to a broader effort by Pi to turn internal tools into services for outside customers. Rather than limiting its products to users already inside its own network, the company is trying to use its existing scale in nodes, user accounts, and identity checks as a commercial proposition for third parties.

That approach places Pi in markets where demand for computing resources, user verification, and account authentication is rising. It also reflects a wider push across the technology sector to build AI-related services closer to end users’ devices, particularly where privacy and infrastructure costs are concerns.

Pi’s identity business is likely to draw particular attention because verification systems with broad international reach are in demand across digital finance and online services. The company claims one of the largest globally distributed verified user bases among crypto-linked platforms.

At the same time, the distributed computing element suggests Pi is seeking additional uses for the network of Node operators built around its blockchain project. If external clients begin paying Node operators for AI-related tasks, that would give Pi a new economic model tied to machine resources rather than solely network participation.

Pi said the latest products are intended to make its services useful both inside and outside its own ecosystem, drawing external developers, businesses, and users into closer contact with its network of verified accounts and user-run machines.

SOFIAH NICHOLE SALIVIO

News Editor

Krotos has released its Video to Sound plugin for Adobe Premiere Pro, available to Krotos Studio subscribers.

The launch brings Adobe’s editing software into Krotos’ broader effort to move more sound design work into the edit timeline rather than a separate audio workflow.

Within Premiere Pro, editors can analyse footage, choose the types of sounds to include, and generate synchronised sound effects on the timeline without leaving the project. Users can set in and out points before creating a sound pass, then swap in alternative sounds while keeping them aligned to picture.

The plugin launches with support for ambiences, whooshes, transitions, risers, impacts and cloth effects. These sounds come from Krotos’ royalty-free library of recorded audio rather than being generated from scratch.

The distinction matters because a growing number of post-production tools now use artificial intelligence not only to organise audio assets but also to create them. Krotos uses machine learning to analyse footage and identify relevant sound moments, while the clips placed on the timeline come from professionally recorded effects and ambience recordings.

That positions the product in a part of the editing market where speed is increasingly important, especially for smaller teams, social video producers and editors expected to complete rough sound design without handing a project to a dedicated audio department. For those users, the appeal is less about replacing full post-production sound work and more about cutting the time spent searching libraries, placing clips manually and rebuilding simple sound beds.

Krotos is based in Edinburgh and develops sound design software for film, television and games. Its tools have been used on productions including Game of Thrones, Avengers: Age of Ultron and Stranger Things, according to the company.

The Premiere Pro release also broadens access to a feature Krotos previewed earlier this year. By making it part of Krotos Studio, Krotos is tying the plugin to its existing subscription platform rather than offering it as a separate standalone product.

Workflow shift

The broader shift in post-production has been toward tighter integration between specialist tools and mainstream editing platforms. Video editors are increasingly handling more of the finishing process themselves, particularly on digital-first projects with smaller budgets and faster turnaround times. That has created an opening for software makers that can automate repetitive parts of sound editing while keeping editors inside familiar applications.

Adobe Premiere Pro is one of the most widely used editing platforms across online video, broadcast packages and independent production. A plugin that works directly in that environment lets Krotos target users who may not move projects into dedicated digital audio workstations until late in the process, if at all.

There are limits to what such tools can do. Complex dramatic sound design, detailed Foley work and final mix decisions still depend on human judgment and, in many productions, specialist teams. Even so, software that can create a first pass of matched effects may appeal to editors producing explainers, trailers, branded clips, documentaries and social content, where time pressure often outweighs the need for highly detailed sound construction.

Subscribers can install the Premiere Pro extension through Adobe Marketplace. The plugin is included with Krotos Studio, Pro and Max subscription plans.

Krotos has built its business around simplifying sound design tasks that have traditionally required specialist knowledge or lengthy manual work. Its products are aimed at a range of users, from professional sound teams to creators with less experience in audio post-production.

For Premiere users, the immediate question is whether automated sound placement can produce results useful enough to keep, or at least strong enough to serve as a draft. Krotos is positioning the tool as a way to turn silent footage into a workable soundtrack quickly, with room for later adjustments rather than a finished mix in one step.

Orfeas Boteas, Chief Executive Officer of Krotos, said the product addresses production schedules that leave editors handling more work in less time.

“Video editors are under constant pressure to deliver more content in less time. With Video to Sound for Premiere Pro, users can go from silent footage to a professional soundscape in just a few clicks, without ever leaving their edit. By combining intelligent analysis with real recordings from our sound library, we’re helping editors spend less time searching for sounds and more time creating,” said Orfeas Boteas, Chief Executive Officer of Krotos.

SOFIAH NICHOLE SALIVIO

News Editor

More than half of UK consumers who use AI assistants say the tools help them discover new brands. But only 4% would buy from an AI-recommended brand without checking elsewhere, according to a separate finding from the same survey.

The research, commissioned by CloudNine PR, surveyed 2,564 UK consumers, including 1,989 who said they use AI chat tools such as ChatGPT.

Among those AI users, 52% said the technology makes it easier to find brands they would never have found otherwise. At the same time, 79% said they would still check other sources before trusting a recommendation from an AI assistant.

The figures suggest a gap between discovery and purchase. While 48% said they would consider buying from a brand suggested by AI even if they had not heard of it before, most still said they would verify the suggestion through more established sources.

Search engines were the most common next step, with 46% saying they would look for the brand on Google or another search engine if an AI tool recommended one they did not know.

Online reviews followed closely, with 43% saying they would check them before deciding whether to trust the recommendation. Another 32% said they would visit the brand’s website, while 27% said they would search for the brand on Amazon.

Smaller shares said they would compare the recommended brand with others, check coverage in online publications or look for social media mentions. Just 19% said they would compare it with competing brands, 10% would check online publications and 9% would look for social media references.

The survey also found that trust rises when AI-generated answers are supported by several sources. Six in 10 AI users said they are more likely to trust recommendations backed by multiple sources, including articles, reviews and influencer content.

The finding adds to a broader debate over how AI assistants are changing online discovery while still relying on trust signals established long before generative AI entered mainstream consumer use. Search results, review platforms, marketplace listings and media coverage continue to shape how consumers judge unfamiliar brands.

Uday Radia commented on the findings.

“AI tools like ChatGPT are rapidly becoming a product discovery channel and they’re helping lesser-known brands get noticed – companies that people might never have come across otherwise. Importantly, however, while AI helps you get discovered it’s not enough to drive conversions on its own. If a consumer doesn’t like what they see about you in Google, independent review sites and earned media, they’re unlikely to become customers,” said Uday Radia, Owner, CloudNine PR.

Advertising concerns

The survey also examined how users might respond if advertising becomes more common in AI-generated answers. The question has become more relevant as major AI and search companies test ways to monetise conversational interfaces.

More than half of AI users, or 54%, said they would trust AI recommendations less if companies’ adverts appeared in AI answers. A larger share, 63%, said they would switch to another AI tool if ads started appearing.

Those responses suggest advertising in AI interfaces could carry reputational risks for providers as they try to build new revenue streams. For brands, the data also indicates that visibility through AI alone may not be enough if users become more sceptical once paid placements are introduced.

The issue is already being tested across the market. OpenAI has expanded its ChatGPT advertising pilot to the UK, while Google has widened ad formats in AI-led search experiences. Anthropic has taken a different position, saying it opposes advertising because it could create incentives that conflict with keeping its Claude assistant helpful.

OpenAI has previously said its ad pilot showed “no impact on consumer trust metrics”, according to comments reported by Reuters. Its partnership with adtech company LiveRamp is designed to help advertisers measure whether ChatGPT adverts lead to purchases.

For brands and retailers, the results underline that AI may now sit near the top of the funnel rather than at the point of conversion. Consumers appear willing to use chatbots to surface unknown products, but they still want proof from search, reviews, websites, marketplaces and media coverage before they spend money.

CloudNine PR said the data was collected by market research company TLF Research from UK consumers, with questions focused on product discovery through AI tools and attitudes towards ads appearing in AI responses.