BRIDGE36 PERSPECTIVES / 01
PHARMA / INVESTMENT STRATEGY
AI in pharma:Which companies will still be worth owning?
AI can make pharma more productive without making every pharma business more valuable. An investor playbook for where value moves - and who can keep it.
01 / THE INVESTMENT SHIFT
More pharma output.
Not necessarily more paid work.
A larger pharma market does not protect every business within it. AI can increase the demand for an outcome while reducing the work customers must buy to achieve it.
WHAT BECOMES EASIER
Producing
the work.
Candidate generation, routine content checks and repeatable production tasks.
WHAT CUSTOMERS STILL NEED
Proving and
delivering value.
Clinical evidence, the right qualified capacity, and reliable decisions in complex workflows.
WHAT THE INVESTOR MUST OWN
A position that
gets paid.
Rights, capabilities or customer access that turn those needs into attractive cash returns.
Research on drug development, manufacturing and content-review automation.168
The opportunity is not simply to automate the work. It is to own a valuable role after the work becomes easier. Ask what the customer still needs to buy from this company - not what it can now produce faster.
02 / THE PHARMA TESTS
What will customers still buy -
and why would they buy it from you?
Test the work pharma will still purchase, the reason to choose this company, and the cash it can retain. Apply the same questions to each product, site or service line.
PAID SCOPE
Will pharma buy more of the outcome - but less of your work?
More drug candidates do not automatically mean more outsourced discovery fees. More content does not automatically mean more paid review hours.
Map the product or task that is purchased - not growth in the wider pharma market.
What would change the investment case?
For each material revenue line, test units bought, work brought in-house and work bundled into software. Weight exposure by gross profit, not employee count. Validate new demand with actual purchasing plans, not the number of AI use cases.
CUSTOMER CHOICE
When every supplier has AI, why would pharma still choose this company?
Generating more molecules is not the same as owning a differentiated drug programme. Producing faster is not a lasting advantage if other qualified suppliers can match the result.
Identify what the customer would lose by switching - even after rival suppliers adopt AI.
What would change the investment case?
Compare the offer with credible alternatives: rights to specific programmes, proven product performance, delivery capabilities and specialist expertise. Test whether useful data, customer relationships or lower costs are genuinely hard to reproduce. Qualification requirements can slow switching; they do not make every qualified supplier distinctive.
RETAINED CASH
Will the company keep the gains - or pass them on through lower prices?
AI-enabled drug programmes can earn licence payments. Packaging savings may be passed on through lower prices. Count what remains after customer concessions, AI costs and the investment still needed.
Trace the payment terms and funding needs - not just the gross savings in the AI business case.
What would change the investment case?
For a drug programme, assess which payments are committed, which depend on future success, and who funds development. For manufacturing or services, test new bids and renewal prices against fully loaded costs. Include implementation, validation, ongoing oversight, capital expenditure and working capital before concluding that returns improve.
Research on acquisition diligence, value capture, defensibility and AI deployment in drug development.234910
03 / WHERE VALUE MOVES
AI can strengthen a business,
squeeze its returns - or remove its role.
Drug rights can create a new revenue stream. Factory efficiency can become the price of staying competitive. Routine checks can move into the customer’s own software. We investigate these outcomes through three hypothetical case studies, informed by real-world developments.
DRUG DEVELOPMENT / AI STRENGTHENS THE BUSINESS
AI can lower the cost of reaching a licensing deal -
while the developer shares in the programme’s value.
HOW AI BECOMES INCOME
Fewer cycles
to select candidates.
Lab evidence.
Programme rights.
A programme
a partner will license.
WHY AI STRENGTHENS THIS COMPANY
Consider a developer that owns its drug programmes and experimental data, with a lab team able to test new candidates. It uses AI to design and select molecules, then tests them in the lab. In this case, it reaches a differentiated programme a partner wants to license with fewer research cycles.
The company’s assets and contract terms turn that improvement into a financial benefit. It licenses development and commercialisation rights to a partner, with an upfront payment, milestones and royalties. Spending less to reach that agreement need not reduce what the partner pays: the partner is buying a promising programme, not reimbursing research hours.
AI strengthens the thesis because it lowers the cost of creating something the company can sell - not simply because the company owns rights.
A programme buyers want.
AI and evidence attract a buyer.More molecules alone do not.
More than access to AI.
Data, labs and rights combine.The tool alone is not the edge.
Less cost to reach payment.
Value-based fees retain savings.Net of AI and development costs.
Invest where AI strengthens an advantage others cannot easily copy.
Here, AI works with the developer’s data, labs and programme rights to create a licensable asset at lower cost. The wider test is the same: does AI combine with company-specific strengths to improve or create a distinctive offering - and can the company retain the value? Fund that combination, not access to the tool alone.
What would validate this investment case?
Check whether AI reduces the time and fully loaded cost needed to produce a programme with evidence a licensing partner values. Confirm data-use and programme rights, buyer interest, deal terms, and who funds the remaining development. Faster discovery does not by itself establish better clinical success.
The real-world starting point is an AI-enabled discovery collaboration combining licensed preclinical programmes, a pharmaceutical partner’s development capabilities, and upfront, milestone and royalty payments. The scenario adds specific assumptions about research-cycle savings and the developer’s assets; it is not an account of that transaction’s realised returns.51
PHARMACEUTICAL PACKAGING / AI BECOMES A REQUIREMENT
AI keeps the factory competitive -
but does not, by itself, make it a market winner.
THE COMPETITIVE EFFECT
Better operations.
Same relative position.
WITHOUT ADOPTIONRisk falling behind.
WITH ADOPTIONKeep pace with the market.
WHY ADOPTION PROTECTS RATHER THAN DIFFERENTIATES
Consider a packaging manufacturer whose qualified competitors offer similar products. It introduces AI-assisted inspection and production planning to reduce defects, waste and delays. The technology works, but other suppliers can adopt comparable tools and operating practices.
As rivals catch up, the improvement becomes what customers expect - not a special reason to choose this factory. Without adoption, it risks losing business to suppliers offering better cost and service. With adoption, it can remain competitive, but customers still have comparable alternatives and can press for lower prices.
AI investment can be necessary to keep this company worth owning, without giving it a reason to grow faster than its market.
The product is still bought.
Customers still buy packaging.AI alone does not create demand.
The improvement is repeatable.
Rivals can make the same gains.Adoption keeps the firm in play.
Cash is protected, not assured.
Savings help defend earnings.After pricing and AI costs.
Treat AI adoption as the cost of staying competitive - not proof of market-beating growth.
Include the spending needed to meet the market’s cost and service standards. Compare adoption with the customers and cash flow at risk without it. The project may still pay back; what it does not establish is a durable premium or market-share gain. Any case for faster growth needs a separate reason customers will choose this supplier.
What would change the competitive conclusion?
Compare cost, quality and delivery with competing suppliers after their likely upgrades - not only with the factory’s own starting point. Check adoption and operating costs, customer renewal terms, and the business at risk if the company does not invest. A hard-to-copy product, process or customer advantage would support a different growth case.
A published pharmaceutical-packaging transformation reported 40% shorter lead times and 71% fewer defects from more than 30 digital use cases, including AI. Those observations inform the operating opportunity. Rival convergence and customer price pressure are the conditions assumed in this scenario, not reported outcomes for that site.672
PROMOTIONAL-CONTENT REVIEW / AI REMOVES THE PAID TASK
AI can make the provider more efficient -
and make its routine service unnecessary.
THE CUSTOMER’S ALTERNATIVE
Checks move
into software.
The customer adopts AI to perform routine promotional-content checks inside its own review workflow.
The same customer need can remain - with fewer tasks bought from an outside provider.WHY COST REDUCTION DOES NOT SOLVE THIS THREAT
Consider a provider whose core service is routine checking of pharma promotional content. Its customers introduce AI into their own review workflows and bring those tasks in-house. They still need compliant content, but no longer need to buy the same checks externally.
Automating the service internally does not keep it needed. It makes the provider cheaper at doing work the customer has stopped purchasing. Even where human review remains necessary, it need not be bought from this provider.
This is a threat to the need for the service - not just to the cost of delivering it.
Outsourced tasks can disappear.
Customers bring checks in-house.The outsourced task disappears.
The customer has an alternative.
Customers can get the same result.The provider loses its role.
Lost sales need a new response.
Savings ease the earnings loss.They cannot replace lost demand.
When AI replaces the paid service, efficiency no longer protects the investment.
Routine checks move in-house in this case. Elsewhere, software or a broader offering can deliver the same outcome without the original service. Establish a new role customers will pay for - or reduce exposure. Automating the provider does not restore demand for work that has been replaced or internalised.
Which customer evidence decides the outcome?
Ask which tasks customers plan to retain internally, what they will continue buying externally, and why. Test a revised service through paying contracts - not a general preference for human involvement.
The real-world starting point is software offered to both in-house pharma teams and agencies to automate routine compliance checks. The scenario assumes customers choose to bring those tasks in-house; the product disclosure does not establish measured provider losses or inevitable displacement. It can also strengthen a provider whose remaining service customers value.8
04 / THE OWNERSHIP PLAYBOOK
Grow the advantage.
Defend the earnings.
Or rethink ownership.
Apply the three tests to each business line - not just to its AI programme.
A stronger
thesis
Invest to grow.
Demand holds and the company gains an edge. It keeps the benefit after costs and investment. Fund growth that turns that edge into cash.
A thesis
to defend
Invest to defend.
Demand holds, but rivals can match the gains. Adoption protects viable earnings, not an edge. Fund competitiveness - not assumed share gains.
A thesis that
no longer holds
Reposition or divest.
Demand or returns undermine the existing thesis. Back a viable new offer with paying customers. Otherwise, weigh a sale against holding for cash.
Separate new earnings from earnings that need investment just to be maintained. Judge the holding on the cash left after that investment - not gross AI savings.
THE OWNERSHIP QUESTION
When AI makes the work easier,
what will this company
still be paid to control?
Bridge36 connects strategic choices with embedded delivery - aligning the commercial proposition, operating model and transformation priorities.
Discuss the investment thesisSources & evidence
Public research and disclosures informing the three hypothetical cases. Links below provide the underlying evidence; the scenarios and investment implications are Bridge36 analysis.
Evidence reviewed 16 September 2026.