Half the world’s population, roughly 20% of global healthcare spending, and no two markets running the same rules. Asia’s opportunity and Asia’s difficulty are the same fact. 

In this episode of Pharma Talks, Nataliya Andreychuk spoke with Gwenael Meneux and Bruno Senuci of drcom about AI in regulated content work and the region most global strategies handle badly. 

Gwenael spent 15 years in pharma, including six in commercial roles across Asia, before moving agency-side just over a year ago. Bruno, co-CEO and co-founder, has lived in Asia for more than 30 years — 18 of them in Vietnam, where he’s based in Saigon — and focuses on operations and innovation alongside Gwenael’s more strategic, pharma-facing role. 

What AI Is Actually Doing 

Teams are using it across drafting, adaptation, translation, repurposing, and MLR readiness. But speed isn’t the objective, and “mostly accurate” doesn’t clear the bar in a regulated industry. 

The question isn’t whether AI can create content, it clearly can. The question is whether that content can be trusted in a heavily regulated environment. 

The gap Bruno points to is that AI can’t guarantee scientific accuracy, doesn’t reliably carry regulatory nuance, and often misses the right context or tone for a given market. 

The recurring mistake is treating it as a shortcut. It’s a multiplier of whatever process already exists — which means a weak process gets scaled, not fixed. Without human oversight, quality control, compliance, and risk management all degrade at the same rate the output accelerates. 

Localization Is Where This Shows 

Translation is solvable. Cultural relevance isn’t, and humans stay accountable for trust — particularly in markets with constraints on channels, operating models, and systems. 

Gwenael’s read is that the current phase of AI adoption has few limits in some fields. Drug development is his example: timelines compressing from eight to ten years toward two to three, which he describes as a change of caliber. 

Content production shows the limits more clearly. The value sits in summarizing, aggregating, and consolidating data — substantial help for medical affairs and marketing. Creating emotion, or adding something meaningful on top of clinical data, still requires human input. World models are evolving, but the boundary is currently palpable. 

MLR Doesn’t Get Removed 

Content can be generated fast. It still has to be validated, and that validation is a human review. AI can support and ease decisions without eliminating the step. 

The companies that do well, Gwenael argues, will be the ones building at the intersection — AI-native services where human teams work alongside agents inside the workflow. Every element of the content supply chain, from localization to market-specific regulatory work, requires preparation, curated knowledge bases, and human accountability at each stage. 

Nataliya raised a vision she’d heard at the Reuters US Pharma Forum: marketers speaking to an agent and watching a finished campaign appear in Salesforce Marketing Cloud. Her caution is that the black-box version is the wrong target. Trusting a model whose reasoning you can’t inspect isn’t a workflow. 

Why the Region Can’t Be Clustered 

Medical need varies more inside Asia than most global plans account for. Some of the world’s oldest populations — Japan, Hong Kong, Taiwan — sit alongside some of the youngest in the Philippines, Cambodia, and India. 

That splits the disease burden. Aging markets look closer to Europe, with high cancer prevalence. Much of Southeast Asia still carries substantial communicable and infectious disease load — dengue, tuberculosis, respiratory conditions tied to air quality. “You cannot cluster these markets the same way,” Gwenael says. 

Health and reimbursement systems diverge just as sharply. China runs a near-universal insurance scheme covering close to the whole population, including drugs on the national list. India’s system is highly fragmented with significant out-of-pocket cost. Access to innovative medicines differs enormously as a result, and that difference propagates into targeting, messaging, and the entire communication approach. 

Launch timing adds another axis. Singapore’s early access programs can put it ahead of some European countries for certain innovations. Other markets in the region may see the same oncology or rare disease drugs three to four years later. One aligned global launch plan cannot cover that spread. 

Channels and Systems Fragment Too 

Bruno’s territory. Vietnam has its own platforms. China’s internet is isolated from the rest. Thailand and Japan have distinct online behaviors. Campaign logic built around Western channels doesn’t replicate — tactics adapt market by market or they don’t work. 

The fragmentation runs deeper than channels. Brands may be managed directly or through distributors. Marketing standards, execution models, and tech stacks vary widely. DAM platforms, CRMs, and approval systems — local, global, or in-house — often differ inside the same company, sometimes between HCP and healthcare communication teams. Silos follow, and integrated execution becomes much harder than an org chart suggests. 

Data regulation fragments as well. Europe has the AI Act and GDPR covering most countries; in Asia each country sets its own rules. And Asian markets generally operate on smaller budgets than Europe or the US, which means adapting across more complexity with fewer resources. Bruno’s team has become expert at doing a lot with much less. 

Bruno’s Framework 

Reduce friction across the full workflow, starting with an audit of local market readiness — technology, talent, regulatory stage, execution capability, budget. 

From there, build agile localization pipelines rather than one-size-fits-all content factories, which he believes have largely failed in Asia. Content needs to be easier to adapt, validate, and approve. AI reduces rework and makes drafts more review-ready, but the approval workflow has to change with it, including tiered approval. 

Roadmaps should be market-specific, with tactics and execution left to local teams. They know how to do it, and don’t need global guidance at that resolution. 

Technology stacks that exist at HQ frequently don’t exist locally — or exist and go underused, because local teams lack the budget, headcount, or skills to integrate them. That points to investment in local digital talent: people who understand how local platforms actually work. 

And don’t wait for the perfect model. Pilot, learn fast, fail fast where necessary, build iteratively. In Asia that beats a complex global model. 

Gwenael’s Framework 

His first warning, drawn from China: assuming what works in China works across Asia. The de-averaged view is the starting requirement. 

That said, some workable clusters aren’t geographically obvious. South Korea, Taiwan, Hong Kong, and Singapore share enough to operate together despite being scattered across the map. 

On content factories, he’s seen highly centralized global models closely, and observes a strong correlation between distance from the global decision center and how poorly equipped a market ends up. He supports regional-scale content factories with unified global templates — avoiding duplication and holding a consistent USP while clustering around ten markets rather than governing a hundred. 

His third recommendation concerns where medical budget goes. Global affiliates tend to spend much of it bringing in global KOLs or sending local KOLs to international congresses. With hindsight, he’d redirect part of that toward raising local expert voices. An Australian or US KOL may be scientifically authoritative and still have far less impact than a trusted local voice when the goal is changing practice in that community. 

Talent and Test-and-Learn 

Even large pharma companies that outsource to distributors and vendors invest in a strong local expert team rather than managing everything centrally. The companies Gwenael saw succeed in China were full of local talent. 

Bruno treats test-and-learn as the single most important practice for Asia — more so than in Europe. He arrived in China wanting everything perfectly structured, and it didn’t work. Going in slightly imperfect and iterating beats pursuing perfection before acting. 

Hot or Not 

Nataliya put statements to both guests and asked for a verdict. Several refused to stay short. 

AI will solve the localization problem in pharma. Bruno declines the clean answer: “It will help. It’s not a magic bullet and the answer is not entirely no. It’s in between.” 

If your content works for Europe, it will work for Asia with a good translation. Gwenael: “Absolutely not. Culture is even more important than language. Understanding how emotion works in Asia, in South Asia versus North Asia, is more important than translation.” 

Human oversight of AI in pharma is non-negotiable. Bruno: “Very hot on that principle. Humans trust humans. We are not there yet at trusting machines blindly.” 

Digital channel fragmentation in Asia is a competitive advantage for those who learn to navigate it. Gwenael: “100% hot. It requires a lot of effort and humility being surrounded by the right local experts, and also having the humility to recognise when it is not the right time to enter a particular market.” 

Global pharma teams understand aging markets well enough to lead digital strategy from headquarters. Bruno: “Not. Absolutely not.” Global teams start from a different tech stack and different principles, skewed toward a Western ecosystem — and Asia’s fragmentation requires adaptation headquarters isn’t equipped to lead. 

The biggest barrier to AI adoption in pharma is technology. Gwenael: “Not. It can be a technology barrier in some cases, but the main one is human adoption and usage. It is a wide change management challenge.” 

Final Words 

Both frameworks converge on the same instruction, applied at different scales. AI multiplies what’s already there — process, content, judgment — which makes it useful to organizations that have those things and dangerous to ones that don’t. 

Asia applies the same logic to geography. De-average the region, invest in people who actually understand the market, and accept that a plan good enough to start with beats a plan complete enough to be late.