In a move that underscores the pharmaceutical industry’s accelerating embrace of artificial intelligence, Eli Lilly is poised to sign a $2 billion deal with Hong Kong-listed Insilico Medicine for AI-developed diabetes drugs. The agreement, which includes a $115 million upfront payment, grants Lilly exclusive rights to sell a GLP-1 drug for diabetes and could total over $2 billion if regulatory and sales milestones are met. This isn’t Lilly’s first foray into Chinese AI partnerships – the company announced a $345 million deal with Shanghai-based XtalPi in November – but it represents one of the largest bets yet on AI’s potential to revolutionize drug discovery.
The China Connection: More Than Just Cost Savings
Lilly’s deal highlights a broader trend: pharmaceutical giants are increasingly looking to China not just for manufacturing, but for cutting-edge innovation. According to data from Evaluate, a record number of pharmaceutical companies from outside China licensed drugs made by Chinese businesses in 2025, totaling $5.6 billion in upfront payments. AstraZeneca’s $4.7 billion licensing deal with CSPC Pharmaceuticals for weight-loss and diabetes drugs in January further illustrates this shift.
But what’s driving this trend? Beyond the obvious cost advantages, Chinese AI companies have gained a competitive edge through more efficient architectures and government prioritization of computing-electricity synergy. According to a Financial Times analysis, Chinese AI models like DeepSeek and MiniMax have overtaken US rivals in token consumption since February 2024, with Chinese companies charging $2-3 per million output tokens compared to $15 for Anthropic’s Claude Sonnet 4.5. This cost advantage becomes particularly significant as AI agents – which can require up to 20 million tokens for minor coding tasks compared to 30,000 for chatbots – become more prevalent.
The AI Promise and Peril in Pharma
Lilly’s CFO Lucas Montarce acknowledged at a March conference that while the company is “investing heavily” in AI for research and development, “it will take more time” to get AI drugs from research to clinical testing. The company’s annual report in February added new language warning about “significant risks involved in developing and deploying AI,” noting that investments may not be effective or profitable and that AI could enable new competitors in drug discovery.
This cautious optimism reflects a broader industry reality: while AI promises to accelerate drug discovery and reduce costs, the path from algorithm to approved medication remains fraught with challenges. The deal with Insilico – founded in 2014 at Johns Hopkins University and an early leader in AI drug development before the current AI frenzy – represents a calculated bet on technology that’s still proving itself in clinical settings.
Beyond Pharma: AI’s Workplace Paradox
The pharmaceutical industry’s embrace of AI mirrors a broader corporate trend, but with a twist. While companies like Lilly invest billions in AI for innovation, other sectors are grappling with AI’s impact on their most human processes: hiring and workforce management.
A recent Financial Times investigation reveals that recruiters are creating “AI-free zones” in interviews as candidates increasingly use AI to write resumes, cover letters, and even answer interview questions. Michael Kienle, global vice-president for talent acquisition at L’Or�al, describes how one jobseeker used AI in a video interview, simply repeating answers the bot provided. “The answers didn’t come naturally,” he explains.
This has prompted companies to reassess their approach. L’Or�al has decided to “sanctuarise the interview” as an AI-free zone, while EY has trained over 20,000 interviewers to “stress-test candidates’ thinking” and spot AI-prepared answers. According to February data from HR platform Deel, more than 40% of employers have extended probation periods as they find it harder to assess true skills during the application process.
The Productivity-Trust Dilemma
Herein lies the paradox: companies are simultaneously investing in AI to boost productivity while creating human-only spaces to maintain trust and authenticity. David Brown, chief executive of recruiter Hays Americas, observes that “AI has actually pushed the interview process back to being more human focused” as employers struggle to decipher between AI-enhanced resumes.
This tension extends beyond hiring. A German study by the Weizenbaum Institute found that while 62% of companies now use AI in regular operations (up from 50% a year earlier), the successful implementation depends heavily on involving worker representatives. In companies where management actively involves works councils, employees report significantly less stress from AI intensification. The study reveals that over 80% of companies use AI primarily for efficiency gains, but instead of mass layoffs, most use the time savings to improve product quality (over 80%) or reduce employee workloads (nearly 75%).
The Economic Implications: Who Benefits?
As AI transforms industries from pharmaceuticals to hiring, questions about who benefits from the productivity gains are becoming increasingly urgent. BlackRock CEO Larry Fink warns in his annual shareholder letter that AI risks intensifying wealth inequality by concentrating gains among businesses and investors who finance AI growth. “The massive wealth created over the past several generations flowed mostly to people who already owned financial assets,” Fink writes. “AI threatens to repeat that pattern at an even larger scale.”
This concern is echoed by Vinod Khosla, an early OpenAI investor, who has proposed eliminating income tax on Americans earning less than $100,000 by raising capital gains levies. He argues this could ease the transition as AI accelerates “a shift of wealth and power away from workers.” Khosla predicts that fear of AI taking jobs will be “the single biggest issue” in the 2028 US presidential election cycle.
A Balanced View of AI’s Impact
Interestingly, public concerns about AI may be shifting. A global survey of over 80,000 Anthropic Claude users across 159 countries found that AI hallucinations (mistakes) are the top concern at 27%, surpassing job displacement at 22%. The study, conducted in 70 languages, found that 32% of users reported increased productivity, with more optimism in lower and middle-income countries.
This suggests that while job displacement fears dominate political discourse, practical concerns about AI reliability may be more immediate for users. As companies like Eli Lilly bet billions on AI’s potential, and others create AI-free zones to preserve human judgment, the technology’s impact appears more nuanced than either utopian or dystopian narratives suggest.
The pharmaceutical industry’s massive investments in AI drug discovery, coupled with other sectors’ cautious implementation, paint a picture of a technology that’s simultaneously transformative and problematic. As AI continues to reshape everything from drug development to hiring practices, the challenge for businesses will be balancing innovation with authenticity, efficiency with equity, and technological promise with practical reality.