Recent coverage of the company’s leadership discussions shows that Coca-Cola is entering what executives describe as a new phase focused on influence not pricing power. According to Mi-3, the company is changing its focus from “price to persuasion,” with digital platforms, AI, and in-store execution becoming increasingly important in building demand. This reflects a change in consumer brand behaviour as inflation pressures ease and companies seek new strategies to maintain revenue growth.

That means expanding the role of AI in Coca-Cola’s marketing production and decision-making. The company has already experimented with generative AI in creative campaigns and continues testing how automation can help with content creation, campaign planning, and distribution.

Industry analysis from The Current points out that Coca-Cola has been embedding AI into marketing workflows and scaling its use in creative production and campaign execution. These efforts include using AI tools to generate images, assist with storytelling, and adjust campaigns in channels.

Testing AI in the marketing pipeline

The week’s reporting suggests the company is now testing AI-driven systems that can help automate parts of the advertising process, including drafting scripts or preparing social media content. While these initiatives remain in testing not full rollout, they illustrate how large brands are moving toward more automated marketing pipelines. Instead of relying only on agencies or long creative cycles, companies are exploring ways to shorten the path from concept to campaign.

During the past two years, many consumer goods have firms relied on price increases to offset rising costs. As inflation slows in several markets, analysts say that strategy has limits. Growth increasingly depends on persuading consumers to buy more often or choose higher-margin products. AI offers a way to refine that persuasion at scale, using data to shape messages, target audiences, and adjust campaigns in near real time.

Coca-Cola’s approach fits a wider trend in marketing technology. Generative AI tools have quickly moved from experimental use to regular deployment in large enterprises. According to McKinsey’s 2024 global AI survey, about one-third of organisations already use generative AI in at least one business function, with marketing and sales among the most common areas of adoption. Analysts expect that share to keep rising as companies test automation in creative work and customer engagement.

AI moves upstream in enterprise strategy

What strikes out in Coca-Cola’s case is how the corporation frames AI not only as a cost-saving tool, but also as part of a broader operating shift. By focusing on persuasion, the company signals that AI’s value lies in shaping demand, not improving efficiency. That includes using AI to analyse consumer behaviour, tailor messaging to different markets, and support local teams with adaptable content.

The strategy also reflects a growing tension in the marketing sector. Automation can speed up production and test more campaign ideas, but it also raises questions about creative quality, brand consistency, and the role of human teams. Companies experimenting with AI-generated content must still ensure that messaging aligns with their brand identity and cultural context. For global brands like Coca-Cola, that challenge becomes more complex because campaigns frequently need to work in many regions.

Another factor shaping this transition is the rapid growth of digital advertising channels. As spending shifts toward social platforms, streaming services, and online retail media, the volume of content required has expanded. AI tools offer a way to produce many versions of ads, test different approaches, and adjust messaging based on performance data. This makes automation appealing not only for cost reasons, but also for speed and flexibility.

Coca-Cola’s move reflects a broader pattern: AI adoption is moving upstream in business processes. Early deployments frequently centred on analytics or internal automation. Companies are now applying AI in customer-facing functions like marketing strategy, creative development, and campaign management. That change suggests that AI is becoming part of how companies compete for market share, not how they reduce expenses.

The firm has not indicated that AI will replace creative teams or agencies. Instead, the current direction indicates a hybrid model in which automation handles repetitive or data-heavy tasks while human teams guide brand voice and campaign concepts. Many marketing leaders believe that this blended approach will define the next phase of AI adoption.

Coca-Cola’s emphasis on persuasion over pricing may impact how other consumer brands approach growth in a post-inflation environment. If AI can assist businesses in more precisely shaping demand, it may minimise reliance on price increases or mass-market campaigns.

(Photo by James Yarema)

See also: PepsiCo is using AI to rethink how factories are designed and updated

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and co-located with other leading technology events. Click here for more information.

AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.

The post Coca-Cola turns to AI marketing as price-led growth slows appeared first on AI News.

That pressure is pushing some large companies to test where AI fits inside everyday marketing work. At L’Oréal, AI-generated creative tools are being used to support parts of the digital advertising process, particularly video and visual content. The aim is not to replace human teams, but to reduce friction in a system that demands constant refresh.

The shift offers a useful view into how enterprise AI adoption is unfolding in creative functions, where speed and control matter as much as originality.

Scaling content without scaling production

For a global beauty group, digital advertising is no longer a seasonal exercise. Content is needed continuously across social platforms, ecommerce sites, and regional campaigns, often with small variations in language, format, or visual emphasis.

Traditional production models struggle to keep up. Each new asset typically involves planning, filming, editing, and approvals. AI-generated images and video elements allow you to reuse old content and extend it into new formats without having to start from scratch every time.

At L’Oréal, AI tools are being used to help generate or adapt visual content that fits specific digital channels. This includes polishing footage, modifying formats, and creating versions for different platforms. Human teams continue to monitor creative direction and final output, but AI speeds up the time between idea and delivery.

The practical value is not about producing something altogether new. It is about producing enough usable content to meet the pace of digital advertising.

Why L’Oréal keeps AI under tight creative control

One reason large brands move cautiously with AI in creative work is brand risk. Visual identity, tone, and messaging are tightly regulated, and small inconsistencies can be amplified when content is distributed at scale.

Rather than handing over creative decisions, companies like L’Oréal are using AI as a support layer. AI-generated output is examined, adjusted, and approved using existing workflows. This keeps accountability with internal teams and external agencies, while still gaining efficiency.

This approach reflects a broader pattern in enterprise AI adoption. Tools are being introduced into workflows that already exist, rather than reshaping how decisions are made. In marketing, that often means AI assists with production, not with defining brand voice.

Cost, speed, and repeatability

Digital advertising budgets are under pressure, even for large consumer groups. Media prices fluctuate, platforms change their restrictions, and audiences expect constant updates. AI offers a way to absorb some of that pressure by lowering the marginal cost of producing additional assets.

By reusing footage and applying AI-based enhancements, brands can stretch the value of each shoot. This is especially important in areas where campaigns must be quickly changed, or when local teams want specific assets but lack full-scale production support.

The result is not a dramatic cost cut in one area, but incremental savings across hundreds of minor decisions. Over time, those savings shape how marketing teams plan campaigns and allocate expenditures.

What this says about enterprise AI maturity

L’Oréal’s use of AI-generated creative work is less about experimentation and more about operational fit. The tools are used in situations where output is predictable, quality can be measured, and mistakes may be caught before release.

This mirrors how AI is being adopted across many enterprise functions. Instead of broad, open-ended use, companies are identifying narrow tasks where AI can reliably assist without introducing new risk. In marketing, those tasks often sit between creative concept and final distribution.

The approach also emphasises a key constraint. AI works best in environments with existing data, rules, and review processes. Creative freedom still belongs to people, while AI supports scale.

Implications for marketing teams

For marketing leaders, the lesson is not that AI will replace agencies or internal creatives. It is that production models built for slower cycles are becoming harder to sustain.

Teams are being asked to deliver more content, more often, with tighter budgets and faster turnaround. AI tools offer one way to manage that demand, but only if they fit existing controls and expectations.

This places new demands on governance. Marketing teams need clear rules on where AI can be used, how outputs are reviewed, and who remains accountable for final decisions. Without that structure, efficiency gains can quickly be offset by risk.

What L’Oréal’s approach signals for enterprise AI adoption

What stands out in L’Oréal’s approach is restraint. AI is applied where it reduces friction, not where it reshapes the role of creative teams. That makes it easier to integrate into large organisations with established processes and brand safeguards.

As more enterprises look to AI for productivity gains, similar patterns are emerging. AI becomes part of the workflow, not the headline. Success is measured in time saved and consistency maintained, not in novelty.

For now, AI-generated creative work remains a supporting act in enterprise marketing. Its real impact lies in how quietly it changes the economics of content production, one asset at a time.

(Photo by Helio E. López Vega)

See also: Disney is embedding generative AI into its operating model

Want to learn more about AI and big data from industry leaders? Check outAI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events, click here for more information.

AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.

The post L’Oréal brings AI into everyday digital advertising production appeared first on AI News.

Disney’s agreement with OpenAI shows how a large, IP-heavy organisation is attempting to resolve that tension by putting AI inside its operating system rather than treating it as a side experiment.

Under the deal, Disney becomes both a licensing partner and a major enterprise customer. OpenAI’s video model Sora will be able to generate short, user-prompted videos using a defined set of Disney-owned characters and environments. Separately, Disney will use OpenAI’s APIs to build internal tools and new consumer experiences, including integrations tied to Disney+. The company will also deploy ChatGPT internally for employees.

The mechanics matter more than the spectacle. Disney is not opening its catalogue to unrestricted generation. The licence excludes actor likenesses and voices, limits which assets can be used, and applies safety and age-appropriate controls. In practice, this positions generative AI as a constrained production layer—capable of generating variation and volume, but bounded by governance.

AI inside existing workflows

A consistent failure mode in enterprise AI programmes is separation. Tools live outside the systems where work actually happens, adding steps instead of removing them. Disney’s approach mirrors a more pragmatic pattern: put AI where decisions are already made.

On the consumer side, AI-generated content will surface through Disney+, rather than through a standalone experiment. On the enterprise side, employees gain access to AI through APIs and a standardised assistant, rather than a patchwork of ad hoc tools. This reduces friction and makes AI usage observable and governable.

The implication is organisational. Disney is treating generative AI as a horizontal capability—closer to a platform service than a creative add-on. That framing makes it easier to scale usage across teams without multiplying risk.

Variation without expanding headcount

The Sora licence focuses on short-form content derived from pre-approved assets. That constraint is deliberate. In production environments, much of the cost sits not in ideation but in generating usable variations, reviewing them, and moving them through distribution pipelines.

By allowing prompt-driven generation inside a defined asset set, Disney can reduce the marginal cost of experimentation and fan engagement without increasing manual production or review load. The output is not a finished film. It is a controlled input into marketing, social, and engagement workflows.

This mirrors a broader enterprise pattern: AI earns its place when it shortens the path from intent to usable output, not when it creates standalone artefacts.

APIs over point tools

Beyond content generation, the agreement positions OpenAI’s models as building blocks. Disney plans to use APIs to develop new products and internal tools, rather than relying solely on off-the-shelf interfaces.

This matters because enterprise AI programmes often stall on integration. Teams waste time copying outputs between systems or adapting generic tools to fit internal processes. API-level access allows Disney to embed AI directly into product logic, employee workflows, and existing systems of record.

In effect, AI becomes part of the connective tissue between tools, not another layer employees must learn to work around.

Aligning productivity with incentives

Disney’s $1 billion equity investment in OpenAI is less interesting as a valuation signal than as an operational one. It indicates an expectation that AI usage will be persistent and central, not optional or experimental.

For large organisations, AI investments fail when tooling remains disconnected from economic outcomes. Here, AI touches revenue-facing surfaces (Disney+ engagement), cost structures (content variation and internal productivity), and long-term platform strategy. That alignment increases the likelihood that AI becomes part of standard planning cycles rather than discretionary innovation spend.

Automation that makes scale less fragile

High-volume AI use amplifies small failures. Disney and OpenAI emphasise safeguards around IP, harmful content, and misuse, not as a values statement but as a scaling requirement.

Strong automation around safety and rights management reduces the need for manual intervention and supports consistent enforcement. As with fraud detection or content moderation in other industries, this kind of operational AI does not attract attention when it works—but it makes growth less brittle.

Lessons for enterprise leaders

1. Embed AI where work already happens. Disney targets product and employee workflows, not a separate AI sandbox.
2. Constrain before you scale. Defined asset sets and exclusions make deployment viable in high-liability environments.
3. Use APIs to reduce friction. Integration matters more than model novelty.
4. Tie AI to economics early. Productivity gains stick when they connect to revenue and cost structures.
5. Treat safety as infrastructure. Automation and controls are prerequisites for scale, not afterthoughts.

Disney’s specific assets are unique. The operating pattern is not. Enterprise AI delivers value when it is designed as part of the organisation’s core machinery—governed, integrated, and measured—rather than as a showcase for what models can generate.

(Photo by Héctor Vásquez)

See also: OpenAI targets AI skills gap with new certification standards

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events, click here for more information.

AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.

The post Disney is embedding generative AI into its operating model appeared first on AI News.

Here, we’re talking about a focus on the practical constraints that determine whether AI changes daily work or merely adds another layer of complexity or tooling.

Brand accuracy a repeatable capability

Marketing agencies’ AI treats brand accuracy as something to be engineered. WPP and Stability AI note that off-the-shelf models “don’t come trained on your brand’s visual identity”, so outputs can often look generic. The companies’ remedy is fine-tuning, that is, training models on brand-specific datasets so the model learns the brand playbook, including style, look, and colours. Then, these elements can be reproduced consistently.

WPP’s Argos is a prime example. After fine-tuning a model for the retailer, the team described how the model picked up details beyond the characters, including lighting and subtle shadows used in the brand’s 3D animations. Reproducing these finer details can be where time disappears in production, in the form of re-rendering and several rounds of approvals. When AI outputs start closer to “finished”, teams spend less time correcting and more time shaping narratives and adapting media for different channels.

Cycle time collapses (and calendars change)

WPP and Stability AI point out that traditional 3D animation can be too slow for reactive marketing. After all, cultural moments demand immediate content, not cycles defined in weeks or months. In its Argos case study, WPP trained custom models on two 3D toy characters so the models learned how they look and behave, including details such as proportions and how characters hold objects.

The outcome was “high-quality images…generated in minutes instead of months”.

The accelerated workflow moves rather than removes production bottlenecks. If generating variations becomes fast, then review, compliance, rights management and distribution, become the constraints. Those issues were always there, but the speed and efficiency of AI in this context shows the difference between what’s possible, and systems that have become embedded and accepted into workflows. Agencies that want AI to change daily operations have to redesign the workflow around it, not just add the technology as a new tool.

The “AI front end” becomes essential

WPP and Stability AI call out a “UI problem”, wherecreative teams lose time interfaces to common tools are “disconnected, complex and confusing”, forcing workarounds and constant asset movement between tools. Often, responses are bespoke, brand-specific front ends with complex workflows in the back end..

WPP positions WPP Open as a platform that encodes WPP’s proprietary knowledge into “globally accessible AI agents”, which helps teams plan, produce, create media, and sell. Operational gains come from cleaner handoffs between tools, as work moves from briefs into production, assets into activation, and performance signals back into planning.

Self-serve capability changes agency operations

AI-powered marketing platforms are also becoming client-facing. Operationally, that pushes agencies to concentrate on the parts of the workflow their clients can’t self-serve easily, like designing the brand system, building fine-tunings, and ensuring governance is embedded.

Governance moves from policy to workflow

For AI to be used daily, governance needs to be embedded where work happens. Dentsu describes building “walled gardens”, which are digital spaces where employees can prototype and develop AI-enabled solutions securely, and commercialise the best ideas. This reduces the risk of sensitive data exposure and lets experiments move into production systems.

Planning and insight compress too

The operational impact is not limited to production. Publicis Sapient describes AI-powered content strategy and planning that “transforms months of research into minutes of insight” by combining large language models with contextual knowledge and prompt libraries [PDF]. Research and brief development compress work schedules, so more client work can happen and the agency has faster responses to shifting culture and platform algorithms.

What changes for people

Across these examples, the impact on marketing professionals is one of rebalancing and shifting job descriptions. Less time goes on mechanical drafting, resizing, and versioning, and more time goes on brand stewardship. New operational roles expand, with titles like– model trainer, workflow designer, and AI governance lead.

AI makes the biggest operational difference when agencies use customised models, usable front ends that make adoption (especially by clients) frictionless, and integrated platforms that connect planning, production, and execution.

The headline benefit is speed and scale, but the deeper change is that marketing delivery starts to resemble a software-enabled supply chain, standardised, flexible where it needs to be, and measurable.

(Image source: “Solar Wind Workhorse Marks 20 Years of Science Discoveries” by NASA Goddard Photo and Video is licensed under CC BY 2.0.)

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and co-located with other leading technology events. Click here for more information.

AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.

The post Marketing agencies using AI in workflows serve more clients appeared first on AI News.

Recent reporting shows the retailer using AI to generate new images of real models wearing different outfits, based on existing photoshoots. Models remain involved in the process, including consent and compensation, but AI is used to extend and adapt imagery without repeating production from scratch. The stated aim is to speed up content creation and reduce the need for repeated shoots.

On the surface, the change looks incremental. In practice, it reflects a familiar pattern in enterprise AI adoption, where technology is introduced not to overhaul how a business works, but to remove friction from tasks that repeat at scale.

How Zara uses AI to reduce friction in repeatable retail work

For a global retailer like Zara, imagery is not a creative afterthought. It is a production requirement tied directly to how quickly products can be launched, refreshed, and sold across markets. Each item typically needs multiple visual variations for different regions, digital channels, and campaign cycles. Even when garments change only slightly, the surrounding production work often starts again from scratch.

That repetition creates delays and cost that are easy to overlook precisely because they are routine. AI offers a way to compress those cycles by reusing approved material and generating variations without resetting the entire process.

AI enters the production pipeline

The placement of the technology is as important as the capability itself. Zara is not positioning AI as a separate creative product or asking teams to adopt an entirely new workflow. The tools are being used inside an existing production pipeline, supporting the same outputs with fewer handoffs. That keeps the focus on throughput and coordination rather than experimentation.

This kind of deployment is typical once AI moves beyond pilot stages. Rather than asking organisations to rethink how work is done, the technology is introduced where constraints already exist. The question becomes whether teams can move faster and with less duplication, not whether AI can replace human judgement.

The imagery initiative also sits alongside a broader set of data-driven systems that Zara has built up over time. The retailer has long relied on analytics and machine learning to forecast demand, allocate inventory, and respond quickly to changes in customer behaviour. Those systems depend on fast feedback loops between what customers see, what they buy, and how stock moves through the network.

From that perspective, faster content production supports the wider operation even if it is not framed as a strategic shift. When product imagery can be updated or localised more quickly, it reduces lag between physical inventory, online presentation, and customer response. Each improvement is small, but together they help maintain the pace that fast fashion relies on.

From experimentation to routine use

Notably, the company has avoided framing this move in grand terms. There are no published figures on cost savings or productivity gains, and no claims that AI is transforming the creative function. The scope remains narrow and operational, which limits both risk and expectation.

That restraint is often a sign that AI has moved out of experimentation and into routine use. Once technology becomes part of day-to-day operations, organisations tend to talk about it less, not more. It stops being an innovation story and starts being treated as infrastructure.

There are also constraints that remain visible. The process still relies on human models and creative oversight, and there is no suggestion that AI-generated imagery operates independently. Quality control, brand consistency, and ethical considerations continue to shape how the tools are applied. AI extends existing assets rather than generating content in isolation.

This is consistent with how enterprises typically approach creative automation. Rather than replacing subjective work outright, they target the repeatable components around it. Over time, those changes accumulate and reshape how teams allocate effort, even if the core roles remain intact.

Zara’s use of generative AI does not signal a reinvention of fashion retail. It shows how AI is beginning to touch parts of the organisation that were previously considered manual or difficult to standardise, without changing how the business fundamentally operates.

In large enterprises, that is often how AI adoption becomes durable. It does not arrive through sweeping strategy announcements or dramatic claims. It takes hold through small, practical changes that make everyday work move a little faster — until those changes become hard to imagine doing without.

(Photo by M. Rennim)

See also: Walmart’s AI strategy: Beyond the hype, what’s actually working

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events, click here for more information.

AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.

The post Zara’s use of AI shows how retail workflows are quietly changing appeared first on AI News.

Roblox keeps AI where the work happens

Rather than pushing creators toward separate AI products, Roblox has embedded AI inside Roblox Studio, the environment where creators already build, test, and iterate. In its September 2025 RDC update, Roblox outlined “AI tools and an Assistant” designed to improve creator productivity, with an emphasis on small teams. Its annual economic impact report adds that Studio features such as Avatar Auto-Setup and Assistant already include “new AI capabilities” to “accelerate content creation”.

The language matters—Roblox frames AI in terms of cycle time and output, not abstract claims about transformation or innovation. That framing makes it easier to judge whether the tools are doing their job.

One of the more practical updates focuses on asset creation. Roblox described an AI capability that goes beyond static generation, allowing creators to produce “fully functional objects” from a prompt. The initial rollout covers selected vehicle and weapons categories, returning interactive assets that can be extended inside Studio.

This addresses a common bottleneck where drafting an idea is rarely the slow part; turning it into something that behaves correctly inside a live system is. By narrowing that gap, Roblox reduces the time spent translating concepts into working components.

The company also highlighted language tools delivered through APIs, including Text-to-Speech, Speech-to-Text, and real-time voice chat translation across multiple languages. These features lower the effort required to localise content and reach broader audiences. Similar tooling plays a role in training and support in other industries.

Roblox treats AI as connective tissue between tools

Roblox also put emphasis on how tools connect to one another. Its RDC post describes integrating the Model Context Protocol (MCP) into Studio’s Assistant, allowing creators to coordinate multi-step work across third-party tools that support MCP. Roblox points to practical examples, such as designing a UI in Figma or generating a skybox elsewhere, then importing the result directly into Studio.

This matters because many AI initiatives slow down at the workflow level. Teams spend time copying outputs, fixing formats, or reworking assets that do not quite fit. Orchestration reduces that overhead by turning AI into a bridge between tools, rather than another destination in the process.

Linking productivity to revenue

Roblox ties these workflow gains directly to economics. In its RDC post, the company reported that creators earned over $1 billion through its Developer Exchange programme over the past year, and it set a goal for 10% of gaming content revenue to flow through its ecosystem. It also announced an increased exchange rate so creators “earn 8.5% more” when converting Robux into cash.

The economic impact report makes the connection explicit. Alongside AI upgrades in Studio, Roblox highlights monetisation tools such as price optimisation and regional pricing. Even outside a marketplace model, the takeaway is clear: when AI productivity is paired with a financial lever, teams are more likely to treat new tooling as part of core operations rather than an experiment.

Roblox uses operational AI to scale safety systems

While creative tools attract attention, operational AI often determines whether growth is sustainable. In November 2025, Roblox published a technical post on its PII Classifier, an AI model used to detect attempts to share personal information in chat. Roblox reports handling an average of 6.1 billion chat messages per day, and says the classifier has been in production since late 2024, with a reported 98% recall on an internal test set at a 1% false positive rate.

This is a quieter form of efficiency. Automation at this level reduces the need for manual review and supports consistent policy enforcement, which helps prevent scale from becoming a liability.

What carries across, and what several patterns stand out:

• Put AI where decisions are already made. Roblox focuses on the build-and-review loop, rather than inserting a separate AI step.
• Reduce tool friction early. Orchestration matters because it cuts down on context switching and rework.
• Tie AI to something measurable. Creation speed is linked to monetisation and payout incentives.
• Keep adapting the system. Roblox describes ongoing updates to address new adversarial behaviour in safety models.

Roblox’s tools will not translate directly to every sector. The underlying approach will. AI tends to pay for itself when it shortens the path from intent to usable output, and when that output is clearly connected to real economic value.

(Photo by Oberon Copeland @veryinformed.com)

See also: Mining business learnings for AI deployment

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events, click here for more information.

AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.

The post Roblox brings AI into the Studio to speed up game creation appeared first on AI News.

The main bottleneck in creating immersive spatial computing experiences – whether for consumer gaming, industrial digital twins, or employee training simulations – has long been the labour-intensive nature of 3D modelling. The production of an interactive environment typically requires teams of specialised artists working for weeks.

WorldGen, according to a new technical report from Meta’s Reality Labs, is capable of generating traversable and interactive 3D worlds from a single text prompt in approximately five minutes.

While the technology is currently research-grade, the WorldGen architecture addresses specific pain points that have prevented generative AI from being useful in professional workflows: functional interactivity, engine compatibility, and editorial control.

Generative AI environments become truly interactive 3D worlds

The primary failing of many existing text-to-3D models is that they prioritise visual fidelity over function. Approaches such as gaussian splatting create photorealistic scenes that look impressive in a video but often lack the underlying physical structure required for a user to interact with the environment. Assets lacking collision data or ramp physics hold little-to-no value for simulation or gaming.

WorldGen diverges from this path by prioritising “traversability”. The system generates a navigation mesh (navmesh) – a simplified polygon mesh that defines walkable surfaces – alongside the visual geometry. This ensures that a prompt such as “medieval village” produces not just a collection of houses, but a spatially-coherent layout where streets are clear of obstructions and open spaces are accessible.

For enterprises, this distinction is vital. A digital twin of a factory floor or a safety training simulation for hazardous environments requires valid physics and navigation data.

Meta’s approach ensures the output is “game engine-ready,” meaning the assets can be exported directly into standard platforms like Unity or Unreal Engine. This compatibility allows technical teams to integrate generative workflows into existing pipelines without needing specialised rendering hardware that other methods, such as radiance fields, often demand.

The four-stage production line of WorldGen

Meta’s researchers have structured WorldGen as a modular AI pipeline that mirrors traditional development workflows for creating 3D worlds.

The process begins with scene planning. A LLM acts as a structural engineer, parsing the user’s text prompt to generate a logical layout. It determines the placement of key structures and terrain features, producing a “blockout” – a rough 3D sketch – that guarantees the scene makes physical sense.

The subsequent “scene reconstruction” phase builds the initial geometry. The system conditions the generation on the navmesh, ensuring that as the AI “hallucinates” details, it does not inadvertently place a boulder in a doorway or block a fire exit path.

“Scene decomposition,” the third stage, is perhaps the most relevant for operational flexibility. The system uses a method called AutoPartGen to identify and separate individual objects within the scene—distinguishing a tree from the ground, or a crate from a warehouse floor.

In many “single-shot” generative models, the scene is a single fused lump of geometry. By separating components, WorldGen allows human editors to move, delete, or modify specific assets post-generation without breaking the entire world.

For the last step, “scene enhancement” polishes the assets. The system generates high-resolution textures and refines the geometry of individual objects to ensure visual quality holds up when close.

Operational realism of using generative AI to create 3D worlds

Implementing such technology requires an assessment of current infrastructure. WorldGen’s outputs are standard textured meshes. This choice avoids the vendor lock-in associated with proprietary rendering techniques. It means that a logistics firm building a VR training module could theoretically use this tool to prototype layouts rapidly, then hand them over to human developers for refinement.

Creating a fully textured, navigable scene takes roughly five minutes on sufficient hardware. For studios or departments accustomed to multi-day turnaround times for basic environment blocking, this efficiency gain is quite literally world-changing.

However, the technology does have limitations. The current iteration relies on generating a single reference view, which restricts the scale of the worlds it can produce. It cannot yet natively generate sprawling open worlds spanning kilometres without stitching multiple regions together, which risks visual inconsistencies.

The system also currently represents each object independently without reuse, which could lead to memory inefficiencies in very large scenes compared to hand-optimised assets where a single chair model is repeated fifty times. Future iterations aim to address larger world sizes and lower latency.

Comparing WorldGen against other emerging technologies

Evaluating this approach against other emerging AI technologies for creating 3D worlds offers clarity. World Labs, a competitor in the space, employs a system called Marble that uses Gaussian splats to achieve high photorealism. While visually striking, these splat-based scenes often degrade in quality when the camera moves away from the centre and can drop in fidelity just 3-5 metres from the viewpoint.

Meta’s choice to output mesh-based geometry positions WorldGen as a tool for functional application development rather than just visual content creation. It supports physics, collisions, and navigation natively—features that are non-negotiable for interactive software. Consequently, WorldGen can generate scenes spanning 50×50 metres that maintain geometric integrity throughout.

For leaders in the technology and creative sectors, the arrival of systems like WorldGen brings exciting new possibilities. Organisations should audit their current 3D workflows to identify where “blockout” and prototyping absorb the most resources. Generative tools are best deployed here to accelerate iteration, rather than attempting to replace final-quality production immediately.

Concurrently, technical artists and level designers will need to transition from placing every vertex manually to prompting and curating AI outputs. Training programmes should focus on “prompt engineering for spatial layout” and editing AI-generated assets for 3D worlds. Finally, while the output is standard, the generation process requires plenty of compute. Assessing on-premise versus cloud rendering capabilities will be necessary for adoption.

Generative 3D serves best as a force multiplier for structural layout and asset population rather than a total replacement for human creativity. By automating the foundational work of building a world, enterprise teams can focus their budgets on the interactions and logic that drive business value.

See also: How the Royal Navy is using AI to cut its recruitment workload

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events including the Cyber Security Expo. Click here for more information.

AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.

The post WorldGen: Meta reveals generative AI for interactive 3D worlds appeared first on AI News.

Built upon the powerful Qwen3-Omni intelligence and trained using a massive dataset with tens of millions of hours of speech data, this isn’t just another AI speech recognition model. The team says it’s designed to deliver highly accurate performance, even when faced with tricky acoustic environments or complex language patterns.

So, how does it stack up against the competition? The performance data, from tests conducted in August 2025, suggests it’s rather impressive.

On a public test for standard Chinese, Qwen3-ASR-Flash achieved an error rate of just 3.97 percent, leaving competitors like Gemini-2.5-Pro (8.98%) and GPT4o-Transcribe (15.72%) trailing in its wake and showing promise for more competitive AI speech transcription tools.

Qwen3-ASR-Flash also proved adept at handling Chinese accents, with an error rate of 3.48 percent. In English, it scored a competitive 3.81 percent, again comfortably beating Gemini’s 7.63 percent and GPT4o’s 8.45 percent.

But where it really turns heads is in a notoriously tricky area: transcribing music. 

When tasked with recognising lyrics from songs, Qwen3-ASR-Flash posted an error rate of just 4.51 percent, which is far better than its rivals. This ability to understand music was confirmed in internal tests on full songs, where it scored a 9.96 percent error rate; a huge improvement over the 32.79 percent from Gemini-2.5-Pro and 58.59 percent from GPT4o-Transcribe.

Beyond its impressive accuracy, the model brings some innovative features to the table for next-generation AI transcription tools. One of the biggest game-changers is its flexible contextual biasing.

Forget the days of painstakingly formatting keyword lists, this system lets users feed the model background text in virtually any format to get customised results. You can provide a simple list of keywords, entire documents, or even a messy mix of both. 

This process eliminates any need for complex preprocessing of contextual information. The model is smart enough to use the context to sharpen its accuracy; yet its general performance is hardly affected even if the text you provide is completely irrelevant.

It’s clear Alibaba’s ambition for this AI model is to become a global speech transcription tool. The service delivers accurate transcription from a single model covering 11 languages, complete with numerous dialects and accents.

The support for Chinese is especially deep, covering Mandarin in addition to major dialects like Cantonese, Sichuanese, Minnan (Hokkien), and Wu.

For English speakers, it handles British, American, and other regional accents. The impressive roster of other supported languages includes French, German, Spanish, Italian, Portuguese, Russian, Japanese, Korean, and Arabic.

To round it all out, the model can precisely identify which of the 11 languages is being spoken and is adept at rejecting non-speech segments like silence or background noise, ensuring cleaner output than past AI speech transcription tools.

See also: Siddhartha Choudhury, Booking.com: Fighting online fraud with AI

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events, click here for more information.

AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.

The post Alibaba’s new Qwen model to supercharge AI transcription tools appeared first on AI News.

Ever watched an AI-generated video and felt like something was missing? The visuals might be stunning, but they often have an eerie silence that breaks the spell. In the film industry, the sound that fills that silence – the rustle of leaves, the clap of thunder, the clink of a glass – is called Foley art, and it’s a painstaking craft performed by experts.

Matching that level of detail is a huge challenge for AI. For years, automated systems have struggled to create believable sounds for videos.

How is Tencent solving the AI-generated audio for video problem?

One of the biggest reasons video-to-audio (V2A) models often fell short in the sound department was what the researchers call “modality imbalance”. Essentially, the AI was listening more to the text prompts it was given than it was watching the actual video.

For instance, if you gave a model a video of a busy beach with people walking and seagulls flying, but the text prompt only said “the sound of ocean waves,” you’d likely just get the sound of waves. The AI would completely ignore the footsteps in the sand and the calls of the birds, making the scene feel lifeless.

On top of that, the quality of the audio was often subpar, and there simply wasn’t enough high-quality video with sound to train the models effectively.

Tencent’s Hunyuan team tackled these problems from three different angles:

1. Tencent realised the AI needed a better education, so they built a massive, 100,000-hour library of video, audio, and text descriptions for it to learn from. They created an automated pipeline that filtered out low-quality content from the internet, getting rid of clips with long silences or compressed, fuzzy audio, ensuring the AI learned from the best possible material.

2. They designed a smarter architecture for the AI. Think of it like teaching the model to properly multitask. The system first pays incredibly close attention to the visual-audio link to get the timing just right—like matching the thump of a footstep to the exact moment a shoe hits the pavement. Once it has that timing locked down, it then incorporates the text prompt to understand the overall mood and context of the scene. This dual approach ensures the specific details of the video are never overlooked.

3. To guarantee the sound was high-quality, they used a training strategy called Representation Alignment (REPA). This is like having an expert audio engineer constantly looking over the AI’s shoulder during its training. It compares the AI’s work to features from a pre-trained, professional-grade audio model to guide it towards producing cleaner, richer, and more stable sound.

The results speak sound for themselves

When Tencent tested Hunyuan Video-Foley against other leading AI models, the audio results were clear. It wasn’t just that the computer-based metrics were better; human listeners consistently rated its output as higher quality, better matched to the video, and more accurately timed.

Across the board, the AI delivered improvements in making the sound match the on-screen action, both in terms of content and timing. The results across multiple evaluation datasets support this:

Tencent’s work helps to close the gap between silent AI videos and an immersive viewing experience with quality audio. It’s bringing the magic of Foley art to the world of automated content creation, which could be a powerful capability for filmmakers, animators, and creators everywhere.

See also: Google Vids gets AI avatars and image-to-video tools

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events, click here for more information.

AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.

The post Tencent Hunyuan Video-Foley brings lifelike audio to AI video appeared first on AI News.

Between wrestling with complicated software, finding someone willing to be on camera, and then spending hours editing out all the “ums” and “ahs,” video production often feels more trouble than it’s worth. Google is aiming to change that narrative with Vids.

So far, it seems to be finding its audience. Google announced that Vids has already rocketed past one million monthly active users, a clear sign that teams are crying out for simpler ways to bring their ideas to life with video.

Your photos now move, and avatars do the talking

Among the latest additions is the ability to turn static images into motion pictures. Imagine you’ve got a great photo of a new product but need something more engaging for a social media post or presentation. You can now upload that picture to Vids, type a quick prompt describing what you want to happen, and Google’s Veo AI will turn it into an eight-second animated clip, complete with sound. It’s a simple way to create eye-catching, brand-aligned content in minutes.

For anyone who dreads being on camera, the new AI avatars will be a welcome relief. This feature lets you produce a polished video without ever stepping in front of a lens. You write your script, choose from a selection of digital presenters, and the AI handles the delivery. It’s perfect for creating consistent training guides, product demos, or team updates without worrying about lighting, background noise, or re-recording twenty takes to get it right.

Google is also tackling the tedious task of editing. A new automatic transcript trimming tool listens to your recordings and, with a few clicks, snips out all the filler words and awkward silences. Speaking from plenty of experience, that will be a huge time-saver.

Building on this, the company confirmed that familiar tools from Google Meet – like noise cancellation, custom backgrounds, and appearance filters – are set to arrive next month. Google Vids will also soon support portrait and square formats, making it much easier to create content for different platforms.

Getting started with Google Vids

With these new tools, Google is trying to make video creation as routine as building a slide deck.

The company is broadening access to Google Vids, making it available to more Workspace customers on business and education plans. Better yet, a basic version of the Vids editor is now completely free for all consumers, offering a range of templates to help you create anything from a tutorial to a party invitation.

To get everyone up to speed, Google has also launched a new “Vids on Vids” instructional series. The video guides walk you through the entire process, demonstrates the best features, and offers practical tips to help you create professional-looking content quickly.

Real businesses are seeing the benefit

Companies are already putting Vids to work. At Mercer International, a global manufacturing firm, it’s being used for employee safety training.

Alistair Skey, CIO of Mercer International, said: “Google Vids has given us the ability to create safety content, developed and curated by our organisation rather than having to go to market to hire very expensive resources to produce that for us.”

It’s also a story of speed and scale. Forest Donovan from the data platform Fullstory was impressed by the efficiency gains. “The amount of [high gloss] content we can create in a matter of hours versus what would normally take weeks has been astounding,” he said.

By embedding these powerful yet simple AI tools directly into its Workspace suite, Google is making the case that video is no longer the exclusive domain of specialist creative teams. It’s becoming a fundamental tool for everyday communication, and these updates just made it accessible to everyone.

See also: Google Cloud unveils AI ally for security teams

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is part of TechEx and is co-located with other leading technology events, click here for more information.

AI News is powered by TechForge Media. Explore other upcoming enterprise technology events and webinars here.

The post Google Vids gets AI avatars and image-to-video tools appeared first on AI News.

The new Hunyuan3D-PolyGen model is Tencent’s first attempt at delivering what they’re calling “art-grade” 3D generation, specifically built for the professionals who craft the digital worlds we play in.

Creating high-quality 3D assets has always been a bottleneck for game developers, with artists spending countless hours perfecting wireframes and wrestling with complex geometry. Tencent reckons they’ve found a way to tackle these headaches head-on, potentially transforming how studios approach asset creation entirely.

Levelling up generating 3D assets

The secret sauce behind Hunyuan3D-PolyGen lies in what Tencent calls BPT technology. In layman’s terms, it means they’ve figured out how to compress massive amounts of 3D data without losing the detail that matters. In practice, that means it’s possible to generate 3D assets with tens of thousands of polygons that actually look professional enough to ship in a commercial game.

What’s particularly clever is how the system handles both triangular and quadrilateral faces. If you’ve ever tried to move 3D assets between different software packages, you’ll know why this matters. Different engines and tools have their preferences, and compatibility issues have historically been a nightmare for studios trying to streamline their workflows.

According to technical documentation, the system utilises an autoregressive mesh generation framework that performs spatial inference through explicit and discrete vertex and patch modelling. This approach ensures the production of high-quality 3D models that meet stringent artistic specifications demanded by commercial game development.

Hunyuan3D-PolyGen works through what’s essentially a three-step dance. First, it takes existing 3D meshes and converts them into a language the AI can understand.

Using point cloud data as a starting point, the system then generates new mesh instructions using techniques borrowed from natural language processing. It’s like teaching the AI to speak in 3D geometry, predicting what should come next based on what it’s already created.

Finally, the system translates these instructions back into actual 3D meshes, complete with all the vertices and faces that make up the final model. The whole process maintains geometric integrity while producing results that would make any technical artist nod in approval.

Tencent isn’t just talking about theoretical improvements that fall apart when tested in real studios; they’ve put this technology to work in their own game development studios. The results? Artists claim to report efficiency gains of over 70 percent.

The system has been baked into Tencent’s Hunyuan 3D AI creation engine and is already running across multiple game development pipelines. This means it’s being used to create actual 3D game assets that players will eventually interact with.

Teaching AI to think like an artist

One of the most impressive aspects of Hunyuan3D-PolyGen is how Tencent has approached quality control. They’ve developed a reinforcement learning system that essentially teaches the AI to recognise good work from bad work, much like how a mentor might guide a junior artist.

The system learns from feedback, gradually improving its ability to generate 3D assets that meet professional standards. This means fewer duds and more usable results straight out of the box. For studios already stretched thin on resources, this kind of reliability could be transformative.

The gaming industry has been grappling with a fundamental problem for years. While AI has made impressive strides in generating 3D models, most of the output has been, quite frankly, not good enough for commercial use. The gap between “looks impressive in a demo” and “ready for a AAA game” has been enormous.

Tencent’s approach with Hunyuan3D-PolyGen feels different because it’s clearly been designed by people who understand what actual game development looks like. Instead of chasing flashy demonstrations, they’ve focused on solving real workflow problems that have been frustrating developers for decades.

As development costs continue to spiral and timelines get ever more compressed, tools that can accelerate asset creation without compromising quality become incredibly valuable.

The release of Hunyuan3D-PolyGen hints at a future where the relationship between human creativity and AI assistance becomes far more nuanced. Rather than replacing artists, this technology appears designed to handle the grunt work of creating 3D assets, freeing up talented creators to focus on the conceptual and creative challenges that humans excel at.

This represents a mature approach to AI integration in creative industries. Instead of the usual “AI will replace everyone” narrative, we’re seeing tools that augment human capabilities rather than attempt to replicate them entirely. The 70 percent efficiency improvement reported by Tencent’s teams suggests this philosophy is working in practice.

The broader implications are fascinating to consider. As these systems become more sophisticated and reliable, we might see a fundamental shift in how game development studios are structured and how projects are scoped. The technology could democratise high-quality asset creation, potentially allowing smaller studios to compete with larger operations that traditionally had resource advantages.

The success of Hunyuan3D-PolyGen could well encourage other major players to accelerate their own AI-assisted creative tools beyond generating 3D assets, potentially leading to a new wave of productivity improvements across the industry. For game developers who’ve been watching AI developments with a mixture of excitement and scepticism, this might be the moment when the technology finally delivers on its promises.

See also: UK and Singapore form alliance to guide AI in finance

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with other leading events including Intelligent Automation Conference, BlockX, Digital Transformation Week, and Cyber Security & Cloud Expo.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

The post Tencent Hunyuan3D-PolyGen: A model for ‘art-grade’ 3D assets appeared first on AI News.

The interactive video generated by Odyssey’s AI model responds to inputs in real-time. You can interact with it using your keyboard, phone, controller, or eventually even voice commands. The folks at Odyssey are billing it as an “early version of the Holodeck.”

The underlying AI can generate realistic-looking video frames every 40 milliseconds. That means when you press a button or make a gesture, the video responds almost instantly—creating the illusion that you’re actually influencing this digital world.

“The experience today feels like exploring a glitchy dream—raw, unstable, but undeniably new,” according to Odyssey. We’re not talking about polished, AAA-game quality visuals here, at least not yet.

Not your standard video tech

Let’s get a bit technical for a moment. What makes this AI-generated interactive video tech different from, say, a standard video game or CGI? It all comes down to something Odyssey calls a “world model.”

Unlike traditional video models that generate entire clips in one go, world models work frame-by-frame to predict what should come next based on the current state and any user inputs. It’s similar to how large language models predict the next word in a sequence, but infinitely more complex because we’re talking about high-resolution video frames rather than words.

“A world model is, at its core, an action-conditioned dynamics model,” as Odyssey puts it. Each time you interact, the model takes the current state, your action, and the history of what’s happened, then generates the next video frame accordingly.

The result is something that feels more organic and unpredictable than a traditional game. There’s no pre-programmed logic saying “if a player does X, then Y happens”—instead, the AI is making its best guess at what should happen next based on what it’s learned from watching countless videos.

Odyssey tackles historic challenges with AI-generated video

Building something like this isn’t exactly a walk in the park. One of the biggest hurdles with AI-generated interactive video is keeping it stable over time. When you’re generating each frame based on previous ones, small errors can compound quickly (a phenomenon AI researchers call “drift.”)

To tackle this, Odyssey has used what they term a “narrow distribution model”—essentially pre-training their AI on general video footage, then fine-tuning it on a smaller set of environments. This trade-off means less variety but better stability so everything doesn’t become a bizarre mess.

The company says they’re already making “fast progress” on their next-gen model, which apparently shows “a richer range of pixels, dynamics, and actions.”

Running all this fancy AI tech in real-time isn’t cheap. Currently, the infrastructure powering this experience costs between £0.80-£1.60 (1-2) per user-hour, relying on clusters of H100 GPUs scattered across the US and EU.

That might sound expensive for streaming video, but it’s remarkably cheap compared to producing traditional game or film content. And Odyssey expects these costs to tumble further as models become more efficient.

Interactive video: The next storytelling medium?

Throughout history, new technologies have given birth to new forms of storytelling—from cave paintings to books, photography, radio, film, and video games. Odyssey believes AI-generated interactive video is the next step in this evolution.

If they’re right, we might be looking at the prototype of something that will transform entertainment, education, advertising, and more. Imagine training videos where you can practice the skills being taught, or travel experiences where you can explore destinations from your sofa.

The research preview available now is obviously just a small step towards this vision and more of a proof of concept than a finished product. However, it’s an intriguing glimpse at what might be possible when AI-generated worlds become interactive playgrounds rather than just passive experiences.

You can give the research preview a try here.

See also: Telegram and xAI forge Grok AI deal

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with other leading events including Intelligent Automation Conference, BlockX, Digital Transformation Week, and Cyber Security & Cloud Expo.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

The post Odyssey’s AI model transforms video into interactive worlds appeared first on AI News.