Pharma’s operational imperatives for 2026: what next for AI, data and connected information delivery?

Current Edition

Pharma’s operational imperatives for 2026: what next for AI, data and connected information delivery?

Sue Tabbitt elicits some predictions for the year ahead from industry experts.

AI will inevitably feature heavily again, but in what context – and where else should companies be focusing their efforts in 2026?

Where and how will AI make a difference?

For pharma leaders blazing a trail with AI, proofs of concept (PoCs) are set to give way to enterprise-scale implementation in 2026. One opportunity for AI to add next-level value in a drug safety context is in enabling predictive pharmacovigilance (PV), according to Lucinda Smith, chief safety product officer at ArisGlobal. AI will also help to hone early signal detection, boosting the impact of insights and evidence generated from real-world data, including data from electronic health records, wearables and so on. That’s as long as companies are geared up for that this might entail:

“Among the challenges that departments face are sourcing the right technology skills, building AI literacy and honing governance around AI,” she notes. “Increasingly PV team members will need a balance of both PV and AI skills.”

^{Lucinda Smith, chief safety product officer at ArisGlobal}

Jean Redmond, chief operating officer at Biologit, which specialises in AI-powered literature monitoring platform for drug safety surveillance, believes regulators have a significant role to play in guiding pharma companies in their adaptation to using AI. “Where 2025 saw draft guidance being released and initial AI programmes being adopted by regulators, this year we can expect to see regulatory authorities drive further frameworks, guidance and expectations for the compliant use of AI, giving organisations the confidence to move from pilot projects into production,” she says.

Smashing silos once and for all

Across late-stage R&D operations, function-specific solutions abound – spanning RIM, PLM, QMS and various workflow tools. This in turn has created integration challenges, which must be addressed in 2026 as orchestration options advance.

“The differentiator in 2026 will be the ability to see the end-to-end work “graph” (products, SKUs, markets, tasks, owners); apply consistent, codified business rules; and coordinate execution across regulatory, manufacturing, quality, supply and commercial,” says Megha Sinha, managing partner and CEO at Kamet Consulting Group, which advises on advanced product lifecycle management spanning multiple functions.

“Instead of treating a re-brand, site transfer or marketing authorisation holder change as dozens of disconnected projects, for instance, organisations will move towards a single orchestration layer that sequences tasks, manages dependencies and continuously re-plans as constraints change. AI will be critical, but as an engine within a work-orchestration fabric,” she notes.

^{Megha Sinha, managing partner and CEO at Kamet Consulting Group}

With fewer boundaries, workload management could be reviewed on a grander scale, says John Cogan, chief operating officer at Qinecsa Solutions, which specialises in pharmacovigilance optimisation. He advocates that life sciences organisations, in a PV context in particular, should “stand back from their global end-to-end operating models, and do a full reanalysis on how their operations need to look beyond 2030,” he says. But this will mean breaking down entrenched structural silos, he acknowledges, which could prove more difficult than any technical integration.

Significant data progress is also needed

How companies manage data will determine the potential for inter-departmental fluidity and process agility. So it is encouraging that structured data now looks set to evolve in earnest from a compliance requirement to strategic infrastructure – driven by EMA’s Network Data Strategy; agreed specifications under IDMP and SPOR; and the fact that AI/automation capabilities depend on high-quality, interoperable data foundations.

“The strategic priority in regulatory operations in 2026 will be the full-scale implementation of structured data across regulatory and interconnected functions,” says Remco Munnik, founder and consultant at Arcana Life Science Consulting. “This shift is being driven by the publication of the EU Network Data Strategy and the tangible progress of EMA’s IDMP and SPOR initiatives, now embedded in core regulatory processes such as PLM eAF for product lifecycle management, and ESMP for shortage reporting.

“PMS is set to become the central source of product data across both the entire product lifecycle and the full regulatory network,” Munnik notes. Consequently, the scope will expand beyond authorised medicinal products to include investigational products under evaluation and for all national competent agencies (NCAs). “This shift demands proactive planning from both industry and regulators,” Munnik says, adding that this is ultimately a ‘people and process’ challenge, since the technical standards are now there to be harnessed.

^{Remco Munnik, founder and consultant at Arcana Life Science Consulting}

Keeping patients central to any improvements

As companies progress their ambitions across multiple dimensions, they must not lose sight of the ultimate beneficiary of any planned improvements: patients. This includes consideration of what is communicated to the market.

“This year, as an industry we need to further transform the way we engage with and reach our time-poor target audiences (particularly healthcare providers, but patients as well),” says Michelle Bridenbaker, chief operating officer at Unbiased Science, a medical affairs and scientific communications consultancy.

“Whether it is for disease-state awareness, or to ensure that patients use medications correctly, we are still struggling in the attention economy in Medical Affairs. Although we have made significant progress over the last five years, there is still so much to be done within the sea of mis- and disinformation, the explosion of social media – and now competition with large language models [LLMs] like ChatGPT and CoPilot – which almost half of clinicians now say they are using[1].

“We need to find ways to operate compliantly in the right social media channels, update our codes of practice, and other legal/compliance barriers to define a robust framework for industry to operate in this space,” she urges. “We also need to find a way to partner with or leverage LLMs to ensure our own content is more visible,” she adds. “As it stands, much of this is behind ‘walls’ which the algorithms, and sometimes HCPs, cannot access. As an industry we must overcome these logistical and legal challenges to ensure that robust, unbiased, and compliant content is being served to audiences, wherever they are.”

^{Michelle Bridenbaker, chief operating officer at Unbiased Science}

Sue Tabbitt is a technology and business journalist with 35 years’ experience. She has covered digital transformation in healthcare and life sciences for the last two decades, and is a senior writer at Sarum Life Sciences in the UK.