IBM and L’Oréal: A New Era of AI in Beauty R&D

Reading Time: 13 minutes

Prefer watching instead of reading? Watch the video here. Prefer reading instead? Scroll down for the full text. Prefer listening instead? Scroll up for the audio player.

P.S. The video and audio are in sync, so you can switch between them or control playback as needed. Enjoy Greyhound Standpoint insights in the format that suits you best. Join the conversation on social media using #GreyhoundStandpoint.

---

Earlier this year, IBM and L’Oréal announced that they have teamed up to develop a foundation model for formulations –  a generative artificial intelligence (GenAI) model embedded right into the heart of sustainable beauty R&D. Per the press release, using decades of L’Oréal’s proprietary formulation data and environmental benchmarks, the duo will co-develop a bespoke AI algorithm to aid the design of formulations leveraging environmentally safe raw materials. The stated intent of this collaboration is to redesign the conventional method of product development to meet sustainability goals and performance standards.

Although this alliance has a futuristic stance, as an analyst, I’m forced to ask: how much of this is tech-driven transformation, and how much is clever marketing? This research note examines the strategic implications, breaks down the core of the IBM-L’Oréal partnership, and provides a realistic lens for technology decision-makers to assess this announcement and what it means for their enterprise.

The Context: When AI Meets Eyeliner

Let’s be honest—AI in cosmetics isn’t the first use-case most of us would consider. Maybe this is exactly what bias looks like since most of us in the technology world are men. But look a little closer, and this partnership reveals a growing trend: industry-specific GenAI solutions that go far beyond language generation and start tackling real-world, highly-regulated, science-driven problems.

L’Oréal’s sustainability agenda—called L’Oréal for the Future—sets clear and aggressive goals, including reductions in Scope 3 emissions, increased use of bio-sourced and biodegradable ingredients, and more circular economy principles across the board. These are not “nice to haves”—they’re business imperatives. And with consumer scrutiny at an all-time high, L’Oréal needs to deliver results beyond clever packaging and buzzwords.

What’s notable here is how L’Oréal is moving away from one-off digital pilots or isolated sustainability campaigns. This is an effort to embed technology at the core of its formulation process—a place where legacy systems, manual research, and time-consuming trial-and-error have long ruled the roost. With IBM in the mix, there’s a credible attempt to industrialise AI inside a traditionally craft-based industry.

Enter IBM, bringing its AI and hybrid cloud expertise to co-create a solution that enables this transition without compromising product safety, performance, or time-to-market.

What’s Unique: Co-Developed, Not Repurposed

To clarify—this isn’t just another foundation model being fine-tuned to write marketing copy. IBM and L’Oréal have gone far beyond the typical playbook of adapting a general-purpose AI model for industry use. The plan is to co-develop from the ground up a foundation model, fuelled by L’Oréal’s proprietary formulation data, internal scientific workflows, and regulatory requirements. That co-development approach is not just a footnote; it’s the heart of the story.

The resulting model does more than recommend ingredients—it’s engineered to evaluate biodegradable, low-impact raw materials that meet an ever-growing list of performance and regulatory criteria. These are not hypothetical optimisations. We’re talking about real-world inputs that must pass scrutiny across safety, efficacy, environmental compliance, and consumer acceptance.

Critically, the model is designed to simulate complex chemical interactions and anticipate how ingredient combinations might perform under different conditions. This is crucial to enable R&D teams to experiment virtually with greater speed and confidence, avoiding costly and time-consuming trial-and-error in physical labs.

Just as important is the model’s alignment with L’Oréal’s sustainability goals—particularly Scope 3 emissions reduction. By integrating supplier data, environmental impact scores, and lifecycle analysis, the model doesn’t just chase efficiency; it embeds ESG principles directly into the innovation process. This ensures that sustainability isn’t a post-formulation checkbox—it’s part of the formulation logic.

The speed of innovation is another standout feature. Because the AI can scan, sort, and simulate vast ingredient combinations in seconds, it allows L’Oréal to bring products to market faster. That speed isn’t just a commercial advantage—it’s a competitive weapon in an industry where trends shift overnight and regulatory scrutiny evolves constantly.

We at Greyhound Research believe perhaps the most compelling outcome of this co-development is how it reshapes organisational workflows. The model may acts as a connective layer between scientists, compliance teams, sourcing managers, and sustainability officers—ensuring that decisions made in the lab can be traced, audited, and defended at the board level.

Succinctly put, this is not AI as a bolt-on. It’s AI as an embedded infrastructure—redefining how product innovation happens from the inside out.

The Business Case: Why AI Is a Game-Changer for Cosmetics R&D

The cosmetics industry is undergoing a radical transformation—one driven as much by shifting consumer expectations as by regulatory pressure and environmental urgency. Gen Z and millennial consumers are no longer satisfied with attractive packaging and clever marketing. They want full transparency, ethical sourcing, and concrete sustainability credentials. At the same time, global regulatory bodies are raising the bar on what qualifies as a “sustainable” product, with increasing scrutiny on everything from ingredient sourcing to lifecycle impact.

For a legacy player like L’Oréal, this creates a dual challenge: move fast enough to meet evolving demand while staying compliant in a hyper-regulated environment. Traditional R&D processes—often reliant on siloed systems, lengthy formulation testing, and human-intensive trial and error—simply can’t scale to address this complexity.

This is where foundation models (GenAI) step in as true game-changers. With a well-architected model, L’Oréal gains the ability to radically accelerate its R&D cycles. Instead of months spent on physical prototyping and testing, AI-generated candidate formulations can dramatically compress timelines, allowing scientists to evaluate hundreds of ingredient combinations in a fraction of the time. This not only reduces costs but significantly speeds up innovation.

Regulatory compliance to material safety also becomes a more streamlined process. The AI is trained on domain-specific formulation data curated over decades meeting constantly updated regulatory constraints. This reduces rework, increases first-time-right outcomes, and ensures consistency across geographies.

On the supply chain front, the foundation model will allow for better foresight into ingredient availability and environmental impact. It will cross-reference supplier data, carbon footprints, and transportation constraints to recommend materials that are not only effective but also responsibly sourced. This helps mitigate risk while reinforcing L’Oréal’s sustainability commitments.

Finally, there’s the benefit of institutional knowledge reuse. Decades of formulation insights, testing logs, and experimental outcomes can be mined and embedded into the model, ensuring that past learnings are not lost to turnover or departmental silos. This makes innovation cumulative rather than repetitive.

We at Greyhound Research believe this, in theory, isn’t just about making things faster or cheaper. The foundation model will help L’Oréal to build an entirely new R&D foundation—one that is more agile, accountable, and aligned with the future of both consumer expectations and global compliance frameworks. Of course, it’s still early days, and one ought to have a dose of reality when assessing outcomes and benefits.

Where the Red Flags Are: Is This Real Progress or PR-Driven Posing?

In the past two decades of tracking this industry and working closely with both buy-side and sell-side clients, I have learned one thing: not to get carried away after reading a press release. While the IBM-L’Oréal collaboration has all the makings of a pioneering move, we’ve seen too many well-publicised tech initiatives buckle under the weight of their ambition. It’s essential to cut through the gloss and look at what could derail this otherwise promising model.

Explainability remains a foundational priority. For a model of this sophistication, regulators, auditors, and internal compliance teams will expect detailed, transparent reasoning behind every recommendation, not merely an intuitive interface. Through the use of the foundation model, L’Oréal will need to clearly demonstrate the rationale behind ingredient selection, ensuring transparent logic that confidently meets regulatory requirements across global jurisdictions.

Addressing data bias proactively will ensure the AI aligns closely with future-oriented sustainability objectives. By carefully managing and updating the historical formulation data used for training, L’Oréal will have to ensure the AI promotes innovative, sustainable ingredient combinations rather than reflecting outdated commercial preferences. This forward-thinking approach supports genuine innovation and positive disruption in formulation practices.

Scalability presents an exciting opportunity. If the model is designed for deployment across L’Oréal’s diverse brands, it must account for numerous product lines and global markets, seamlessly adapting to evolving regulatory landscapes in regions such as the EU, India, and the US. Its architecture will have to allow for efficient handling of new formulation challenges, even those outside its initial training data, ensuring ongoing innovation and relevance.

Greyhound Research emphasizes the importance and value of addressing these aspects proactively. Successfully managing these operational and reputational dimensions will significantly enhance credibility and reinforce trust. In today’s environment, demonstrating authentic transparency and genuine sustainability practices can amplify the positive impact and strengthen consumer confidence.

Industry Impact: More Than a Cosmetic Shift

Despite the overall complexity, if there’s one thing that is certain, it is that this collaboration between IBM and L’Oréal isn’t just another tech experiment. If anything, it marks a pivotal moment in how industries are reimagining sustainability at scale. While the immediate impact is visible in beauty and personal care, the ripple effects of this initiative are likely to extend far beyond lipsticks and serums.

IBM’s move into this space signals a much larger trend of technology vendors beginning to carve out vertical offerings tailored for sustainability-focused transformation. With IBM laying the groundwork, expect other tech giants like Salesforce, Oracle, AWS, and Google Cloud to follow suit, bundling AI capabilities that go beyond marketing automation or customer experience. They’ll likely build AI sustainability toolkits that are deeply integrated with operational systems—finance, procurement, manufacturing, and supply chain.

Moreover, we’re starting to see a new class of ESG benchmarks emerge, many of which rely not only on reported metrics but on verifiable, real-time data flows. If IBM’s foundation model helps L’Oréal meet or even exceed these expectations—by identifying better raw materials, reducing Scope 3 emissions, and ensuring regulatory compliance—it could become the model regulators point to when drafting future policy. This could fundamentally change how sustainability is measured, reported, and enforced across industries.

There’s also a profound shift happening in how foundation models are positioned within the enterprise. Until now, much of the enterprise AI excitement has lived downstream—in customer service, copywriting, and basic automation. But IBM and L’Oréal are moving foundation models upstream, embedding it in scientific research, R&D decision-making, and compliance-driven innovation. That shift is not only harder to execute but also far more transformative.

The most compelling signal? This initiative will likely force adjacent industries to reevaluate their own approach. Pharma companies might begin rethinking molecule discovery using similar AI co-development models. Agritech firms may explore how AI can optimise crop inputs with environmental variables in mind. Even aerospace manufacturers, under pressure for net-zero targets, could use this model to rethink materials engineering and lifecycle impact.

In short, the IBM-L’Oréal partnership has opened the door to a much broader redefinition of how foundation models can drive sustainability—not as a bolt-on feature, but as a core business process enabler.

What L’Oréal Stands to Gain

If execution aligns with ambition, L’Oréal stands to benefit in both immediate and long-term ways.

With a foundation model dramatically accelerating the formulation and testing lifecycle, L’Oréal could compress what once took months into just a few weeks. This speed gives the company greater agility to respond to market trends, regulatory shifts, or consumer demands—without sacrificing quality or compliance.

Another key gain is proactive compliance. Embedding regulatory logic into the AI’s formulation process means potential issues can be flagged before human testing even begins. This reduces late-stage reformulations and ensures smoother approvals in diverse jurisdictions. In highly regulated markets like the EU or China, that’s not just nice to have—it’s essential.

L’Oréal also stands to gain from improved internal collaboration. AI breaks down silos between formulation scientists, regulatory teams, and sustainability officers by providing a unified platform for decision-making. Coupled with a sharp reduction in raw material testing cycles, teams can now spend less time validating guesses and more time innovating.

Looking at the long term, the strategic rewards are even greater. Strengthening its ESG credentials will help L’Oréal earn trust with investors, regulators, and customers—especially in an era when ESG metrics are under scrutiny for greenwashing. A system like this, backed by traceable data and measurable impact, gives L’Oréal a real shot at becoming an industry standard-bearer.

There’s also the infrastructure play. The model’s architecture can be extended across brands, geographies, and business units—creating a scalable innovation engine. Rather than building one-off tools for each product line, L’Oréal can build once and deploy many times.

Perhaps most importantly, this partnership makes L’Oréal a first mover. That comes with real strategic weight—setting precedents, shaping regulatory thinking, and defining how AI should operate in beauty and sustainability. In a sea of hesitant incumbents and cautious adopters, that leadership could be priceless.

In short, this isn’t just a win for product development. Done right, it’s an enterprise-wide infrastructure shift that delivers operational, strategic, and reputational returns.

Competitive Landscape: Parallel Movements Worth Watching

While IBM and L’Oréal may have taken the spotlight with this co-developed foundation model initiative, they’re not alone in trying to rewire the relationship between AI and sustainability in the consumer goods sector. Several industry players are already laying the groundwork, albeit through different routes and with varying levels of depth.

Unilever, for instance, has partnered with Microsoft and embraced Azure’s AI cloud platform to optimise operations across its global supply chain. At its Tinsukia facility, the company achieved a staggering 85% reduction in product changeover times and reported a 400% boost in labour productivity. These aren’t just incremental improvements—they represent a shift in operational tempo driven by AI and automation.

Meanwhile, Estée Lauder Companies (ELC) has gone all in on AI-powered trend detection and consumer sentiment analysis. One notable win came from its use of AI to identify the rising popularity of “peach makeup” trends—months before they went mainstream. This early insight translated into faster product rollout and sharper marketing alignment. ELC is also investing in sustainability profiling across its portfolio, using data to assess packaging impact, formulation safety, and long-term ecological footprint.

Procter & Gamble (P&G), another heavyweight, continues to experiment through its LifeLab initiative. AI is helping the company test new ingredient alternatives, evaluate the viability of sustainable packaging materials, and improve the speed of R&D cycles. Although less vocal about domain-specific GenAI, P&G’s moves indicate a long-term interest in embedding AI across product lifecycle stages—from concept to consumer.

That said, we at Greyhound Research would like to point out that none of these players have gone public with a GenAI model that matches the scale, specificity, and co-development depth of the IBM-L’Oréal partnership. And that’s precisely what makes this collaboration a standout—at least for now. It’s not just about who’s using AI; it’s about who’s building it from scratch to solve deeply complex, industry-specific problems.

Boardroom and Investor Implications

Beyond the lab, the tech, and the R&D implications, this collaboration holds significant strategic weight in the boardroom and among institutional investors. L’Oréal isn’t just experimenting with foundation models; it’s signalling to markets that it’s serious about innovation-driven sustainability and that it’s ready to be held accountable for both promises and outcomes.

For investors who are increasingly tying capital allocation to ESG compliance and long-term value creation, this initiative serves as a strategic marker. ESG metrics are no longer a side dish—they’re the main course in portfolio strategies. L’Oréal’s ability to embed AI into its sustainability roadmap and demonstrate verifiable, data-driven progress gives it a distinct edge in attracting capital from ESG-focused funds and asset managers. In particular, if the company can surface reliable real-time Scope 3 emission metrics and show reductions enabled by this AI engine, it could strengthen its weighting on sustainability indices and ETFs.

Reputationally, the stakes are equally high. Public trust around AI is shaky, and greenwashing accusations are no longer confined to small activist corners—they’re boardroom-level concerns. Suppose this partnership can deliver measurable results with clear provenance, L’Oréal gains more than brand lift. It earns the kind of reputational capital that cushions against regulatory pressure, consumer scepticism, and even activist shareholder interventions. Transparency, traceability, and auditability will be key.

Board-level accountability is another lever. ESG and digital transformation have become regular fixtures on board agendas, often discussed in silos. This initiative creates a common thread—tying AI investments directly to ESG targets. CIOs and CSOs (Chief Sustainability Officers) must now collaborate on reporting structures, progress metrics, and cross-functional governance. The foundation model system itself can become a boardroom dashboard—if properly integrated and operationalised.

We at Greyhound Research are of the firm belief that this collaboration could serve as a template for future board engagement around AI. It shifts AI from experimental playgrounds into enterprise-wide strategic initiatives with real shareholder implications. For companies that follow, the message is clear: domain-specific foundation models are no longer optional, and neither is ESG. The ones who crack the code to unify both—transparently, securely, and at scale—will define the next era of corporate leadership.

How CIOs Can Operationalise This Playbook

For CIOs watching this development, the question is no longer “Should we explore this?” but “How do we build something similar—and make it stick?” The IBM-L’Oréal partnership offers a clear blueprint, but successful replication depends on execution, cross-functional alignment, and relentless governance.

The first step is a full audit of your R&D data estate. AI models are only as good as the data they learn from—and if your formulation records are buried in Excel sheets, siloed databases, or handwritten lab notebooks, you’re starting from behind. CIOs must work with R&D, compliance, and procurement teams to build a unified, AI-ready data architecture that’s secure, well-annotated, and accessible across functions.

From there, establish an AI working group with representation from science, sustainability, legal, IT, and product development. Too many AI initiatives fail because they start in isolation. Governance must be built into the foundation, not bolted on later. Define who owns what, how decisions are logged, and where the human-in-the-loop checkpoints live.

Crucially, select partners who understand your industry’s scientific and regulatory depth. You’re not looking for a generalist cloud vendor—you need a collaborator who brings both domain fluency and the technical muscle to co-develop, not just sell, the right solution. The IBM-L’Oréal model works because both parties invested in a true build, not a retrofit.

When it comes to KPIs, go beyond speed and automation. Build a metrics framework that includes traceability, emissions impact, regulatory alignment, model explainability, and user adoption. An AI model that delivers answers but lacks trust or transparency is worse than no model at all. Ensure your dashboards reflect the business’s real definition of success, not just the IT departments.

Finally, foundation models should be elevated from an IT initiative to a board-level agenda. As seen with L’Oréal, this isn’t just a digital transformation effort—it’s an enterprise strategy shift. CIOs should create a reporting cadence that ties AI performance to broader ESG goals, product timelines, and innovation outcomes. This makes foundation models tools of strategic execution, not experimentation.

Done right, this is more than an initiative. It’s a new operating model for innovation.

Final Word: Bold Move, Real Stakes

The IBM-L’Oréal collaboration is more than a partnership—it’s a moment of reckoning for how AI can (and should) serve sustainability. It’s bold, yes. But it’s also untested at this scale and scope. There are real risks here: technological, regulatory, reputational. But there’s also something more important—an opportunity to reshape not just beauty R&D but how innovation is structured across all sectors where regulation, science, and sustainability collide.

If this works—and that’s still a big if—it will become the definitive case study on how to align foundation models with ESG imperatives. It’ll validate the co-development approach, set new standards for AI governance, and elevate AI from an operational tool to a strategic imperative. Other industries will follow. Investors will watch. Regulators might even model future frameworks based on what unfolds here.

This is about reimagining how enterprises solve problems that matter—climate, compliance, consumer trust—and doing it at scale, with accountability. At Greyhound Research, we view this as a live experiment with implications far beyond beauty. We’ll be tracking how the tech evolves, how the metrics hold up, and how decision-makers across industries respond.

Because what’s on the line here isn’t just a cleaner product—it’s a cleaner, smarter, more responsible blueprint for enterprise transformation.