# Claude Science, Anthropic's risky bet ahead of its IPO > On June 30, 2026, Anthropic unveiled Claude Science, a digital workspace built for researchers, preconfigured with more than 60 scientific databases spanning - Author: Maxime Marquette - Publisher: MadMax (https://mad-max.co) - Section: Commentary - Published: 2026-07-06 - Canonical URL: https://mad-max.co/en/article/claude-science-le-pari-risque-d-anthropic-avant-son-entree-en-bourse ## Key takeaways - Claude Science, unveiled June 30, 2026, runs on existing Claude models and offers preconfigured databases such as UniProt, PDB and ChEMBL, available in beta on macOS and Linux for paid subscribers. - Anthropic confidentially filed for an IPO on June 1, 2026, after a Series H round valuing the company at about 965 billion dollars. - The drug program will first target neglected diseases, though Anthropic has not specified which ones or detailed partnerships, and did not respond to reporters' questions. - Researchers quoted in the article stress that AI has not made experiments unnecessary and that no AI-designed drug has yet received FDA approval. - The columnist raises concerns about access gaps for under-resourced labs, dual-use biosecurity risks, and the timing of announcements relative to the IPO. ## Introduction: when an AI lab wants to become a drug company ### A launch that changes the very nature of Anthropic On June 30, 2026, **Anthropic** unveiled **Claude Science**, a digital workspace built for **researchers**, preconfigured with more than **60 scientific databases** spanning genomics, proteomics, structural biology, and computational chemistry (Reuters). The tool is not a new artificial intelligence model, but rather a platform that organizes **scientific** work around Anthropic's existing Claude models, including **Claude Opus 4.8** (TechCrunch). What truly sets this announcement apart is **Anthropic**'s parallel decision to enter drug development itself, a move confirmed by **Eric Kauderer-Abrams**, the company's head of life sciences, who said the effort would first target **neglected diseases**, largely abandoned by the traditional pharmaceutical industry (The Verge). ### Why this precise moment is no coincidence This announcement comes as **Anthropic** confidentially filed for an initial public offering with the SEC on June 1, 2026, after closing a Series H funding round of **65 billion dollars** that values the company at roughly **965 billion dollars** (SmartAsset). An IPO has been floated as early as October 2026, which could potentially make **Anthropic** the first artificial intelligence company to cross the symbolic trillion-dollar valuation mark (IG). Diversifying revenue sources ahead of an IPO under this much scrutiny is far from trivial: it sends a clear signal to investors that the company does not intend to depend solely on chatbot subscriptions or **Claude Code**, its coding tool that was already generating more than 2.5 billion dollars in annualized revenue as of February 2026 (IG). *I find this diversification strategy both brilliant and perilous. Brilliant because it positions Anthropic as far more than a mere chatbot vendor ahead of its IPO. Perilous because drug development is a completely different business, one where technological promises collide brutally with biological reality.* ## What Claude Science actually offers researchers ### An architecture designed to eliminate technical friction The platform runs on a lead coordinating assistant that acts like a project manager, capable of spinning up specialized sub-assistants to divide **scientific** work, somewhat like a supervisor delegating precise tasks to experts (TechCrunch). A separate reviewer agent then checks citations and calculations before any **publication**, a quality-control layer rarely automated to this degree. The tool can run code on a lab's own infrastructure instead of sending **data** to **Anthropic**'s servers, a choice that directly addresses the privacy concerns of **research** institutions and partner pharmaceutical companies (TechCrunch). ### Recognized databases built in natively Among the **databases** listed are **UniProt**, **PDB**, **Ensembl**, **Reactome**, **ClinVar**, and **ChEMBL**, all queryable in natural language without the researcher having to write a custom **data** pipeline for each one (Pondero). The platform can also generate three-dimensional protein structures, genome browser tracks, and chemical representations directly within the work environment. The tool is currently available in beta for **Pro**, **Max**, **Team**, and **Enterprise** subscribers on macOS and Linux, with no Windows support at launch, which for now limits access to a tech-savvy slice of the **scientific** community (Evermx). *I'm impressed by the technical ambition of this multi-agent architecture, but I remain cautious about the absence of Windows support and the requirement of a paid subscription with no free tier. That risks widening the gap between well-funded Western labs and researchers in countries with far fewer resources.* ## The drug discovery program and its blind spots ### Neglected diseases, but few concrete details **Anthropic** has chosen to prioritize **neglected diseases**, the ones the traditional pharmaceutical industry often abandons for lack of sufficient profitability, a direction that could have a genuine humanitarian impact if it materializes (The Verge). Even so, the company has offered very few specifics about which diseases it will target first. Asked about potential partnerships for lab trials, animal testing, **clinical trials**, or manufacturing, **Anthropic** did not respond to reporters' requests for clarification, leaving significant uncertainty hanging over the concrete feasibility of this program (The Verge). ### Aggressive hiring in the life sciences Over the past year, the company has actively recruited biologists and built its own wet labs, an unusual physical infrastructure for an artificial intelligence company accustomed to operating almost exclusively in the digital realm (The Verge). Researcher **Namshik Han** says **Anthropic** has successfully poached several candidates from major pharmaceutical companies and prestigious academic institutions. This massive hiring push shows just how seriously the company is taking this diversification, even though the **scientific** outcome remains highly uncertain at this early stage of the project. *I welcome the intent to tackle neglected diseases, a choice that breaks from the purely commercial logic of the pharmaceutical industry. But I refuse to give in to enthusiasm without concrete proof: without details on targeted diseases or industrial partnerships, this announcement remains, for now, a statement of intent.* ## Pharmacology experts' justified skepticism ### AI hasn't made experiments obsolete yet Researcher **Frank von Delft** summed up the situation with welcome frankness, saying that artificial intelligence models **have not yet managed to make experiments unnecessary**, adding that if **Anthropic** truly wants to develop a drug, the company will have to invest heavily in traditional experimental trials (The Verge). This fundamental **scientific** reality does not vanish simply because a technology company takes an interest in it. Researcher **Matthew Todd**, for his part, notes that the field remains **very far** from seeing an artificial-intelligence-designed drug win regulatory approval for human use, stressing that there is always a **lengthy timeline** required to experimentally demonstrate the safety of a **treatment** (The Verge). ### Human oversight that remains indispensable **Todd** also insists that the drug **discovery** process cannot run autonomously, requiring constant human supervision throughout development, a crucial point that technology companies' marketing communications sometimes tend to downplay (The Verge). To date, no artificial-intelligence-designed drug has yet completed the full path of **clinical trials** through to **FDA** approval. Some candidates developed with the help of artificial intelligence have indeed entered **clinical trials**, but it remains difficult to precisely assess AI's real contribution to these partial successes, or whether these candidates truly outperform drugs developed through conventional methods (The Verge). *I believe this dose of scientific skepticism is exactly what our era needs. Overblown enthusiasm around artificial intelligence in healthcare rarely benefits patients in the short term; it mostly benefits the stock prices of companies that know how to wield the right vocabulary at the right moment.* ## Fierce competition in AI applied to science ### Anthropic joins an already crowded pack **Anthropic** is not the first **Western** technology company to move into this space: **OpenAI**, **Amazon**, and **Google** already have their own tools and platforms dedicated to the life sciences (The Verge). The company also joins a broader race that includes pharmaceutical companies built natively around artificial intelligence, such as **Insilico**, as well as **Isomorphic Labs**, the **Google DeepMind** subsidiary specializing in drug **discovery**. What sets **Anthropic** apart, however, is its peculiar position: the company now sells software to other pharmaceutical companies while potentially becoming their direct competitor in the **discovery** of new **treatments**, an unusual business situation that could create friction with its own customers (The Verge). ### Prestigious clients already on board The company has cited **Novo Nordisk** and the **Allen Institute** as case **studies** among its early clients, suggesting that leading pharmaceutical organizations are already collaborating with multiple artificial intelligence providers simultaneously (TechCrunch). This early adoption by recognized institutions lends the project a degree of initial credibility. **Anthropic** is also funding up to 50 external **research** projects in artificial intelligence applied to science, with grants of up to **30,000 dollars** each, with applications open until July 15, 2026 (TechCrunch). *I see this fierce competition among Western tech giants as a valuable engine of innovation, provided it doesn't turn into a headlong rush where scientific caution is sacrificed on the altar of commercial speed and stock market pressure.* ## The Western strategic stake behind this announcement ### Why the West must dominate this scientific race This announcement fits into a broader context of global technological competition in which the **West** must imperatively keep its lead in **scientific research** applied to artificial intelligence, facing strategic rivals such as **China**, which is investing massively in its own biotechnology capabilities. The ability to accelerate **scientific** **discovery** is not merely a commercial matter, it is a long-term question of technological sovereignty. Companies like **Anthropic**, which apply rigorous safety and **biosecurity** evaluations to their models before any **scientific** deployment, offer a reassuring contrast to less scrupulous players who might neglect these essential safeguards in their own race to innovate (Reuters). ### The risk of a divide between rich and poor labs **Claude Science**'s current business model, restricted to paying subscribers with no free tier, nonetheless risks widening a gap between well-funded **Western** labs and **research** institutions with more modest resources, particularly in developing countries where access to this kind of tool could transform public health research. This question of accessibility deserves particular attention from **Western** policymakers, because a **scientific** innovation that benefits only the wealthiest ends up widening global inequalities rather than reducing them. *I believe the West has every interest in dominating this scientific race, but that dominance must come with a minimum of responsibility toward less-favored countries. A scientific AI that benefits only the richest labs betrays part of its original humanitarian promise.* ## What the tight timing of these multiple announcements reveals ### A communication cadence that raises questions The pace of **Anthropic**'s announcements in recent weeks, from the confidential IPO filing on June 1, to the launch of **Claude Science** on June 30, to the parallel revelations about its drug development program, traces out a communications strategy clearly orchestrated to maximize the attention of potential investors. This synchronization is not inherently suspicious, but it calls for a cautious reading: each announcement must be judged on its actual **scientific** merits, not simply accepted because it comes from a company in the midst of preparing one of the most significant technology IPOs in recent history (VNCMac). ### The question of long-term credibility If **Claude Science** and the drug **discovery** program produce tangible **scientific** **results** in the years to come, this current period will be seen in retrospect as the pivotal moment when **Anthropic** stopped being merely a chatbot company and became a full-fledged scientific player. If results are slow to arrive, the announcement risks being seen as a mere public-relations maneuver ahead of an IPO. Time, not press releases, will settle this essential question for the company's long-term credibility. *I remain deliberately cautious in the face of this avalanche of strategically timed announcements. Recent Silicon Valley history is full of examples where media enthusiasm ran far ahead of real scientific results. The benefit of the doubt must come with constant journalistic vigilance.* ## Historical precedents of AI in pharmacology ### Repeated promises, still modest results The recent history of artificial intelligence applied to drug **discovery** is dotted with ambitious promises that often took longer than expected to materialize. Companies like **Insilico Medicine** have indeed managed to advance drug candidates into **clinical trials**, but none has yet crossed the full finish line to commercial regulatory approval. This reality does not invalidate the technology's potential, but it is a reminder that between an impressive technical demonstration and a drug truly approved for patients lies a considerable gap in time and money that few technology companies fully account for when they make their initial announcements. ### The lack of quality experimental data still holds the sector back **Researchers** **Todd** and **Han** both stress that the lack of high-quality public experimental **data** could considerably slow efforts to develop drugs through artificial intelligence, a structural obstacle that even **Anthropic**'s considerable financial resources will not overcome overnight (The Verge). This fundamental technical limitation deserves repeating every time a new technology company announces its intent to revolutionize pharmacology through artificial intelligence. *I believe these historical precedents should encourage methodical caution rather than cynical dismissal. AI is genuinely progressing in this field, but at a far slower pace than the triumphant press releases of Western technology companies would suggest.* ## Precedents among rival technology giants ### Google DeepMind and Isomorphic Labs lead the way **Google DeepMind**'s subsidiary, **Isomorphic Labs**, has been working for several years on applying artificial intelligence to drug **discovery**, drawing notably on advances from the **AlphaFold** system for protein structure prediction. This accumulated experience gives **Google** a methodological head start that **Anthropic** will have to catch up on quickly. **OpenAI** has also multiplied partnerships with biotechnology companies, betting on a different approach that combines its own generative models with specialized partner labs rather than building internal wet labs as **Anthropic** is now doing. ### A diversity of strategies that ultimately benefits research This plurality of approaches among the major **Western** artificial intelligence labs is, paradoxically, good news for overall **scientific** progress, with each company exploring different angles that could complement rather than needlessly duplicate one another. Competition among these technology giants also accelerates the pace of **scientific** **publications** and methodological innovations, a side benefit rarely highlighted in media coverage of this commercial rivalry. *I see this strategic diversity among Western giants as a collective strength rather than mere commercial fragmentation. As long as these companies remain based in the West and subject to our ethical standards, their rivalry effectively serves the whole of global scientific research.* ## The role of external funding agreements ### An attempt to rally the independent scientific community The grant program launched by **Anthropic**, offering up to 30,000 dollars to 50 selected external projects, explicitly targets doctoral students and postdoctoral **researchers** exploring the frontiers of science across multiple disciplines, with a particular focus on biomedical **research**. Applications will be evaluated and **results** announced before the end of July 2026. This kind of external funding, while modest compared to the company's overall resources, could have a significant multiplier effect by giving talented young **researchers** access to tools normally reserved for large, well-funded institutions. ### A computing partnership that lowers technical barriers Infrastructure partner **Modal** is providing additional computing credits worth up to 2,000 dollars for certain selected grant recipients, complementing **Anthropic**'s offering to reduce technical barriers to entry for independent **researchers**. These funded projects are expected to run between September and December 2026, with the first public **results** expected in early 2027, which will offer a first concrete glimpse of this platform's real value for independent **scientific research**. *I welcome this external funding initiative, which could partially democratize access to scientific tools normally reserved for the best-funded labs. But 30,000 dollars per project remains a modest sum against the real needs of cutting-edge biomedical research.* ## The implications for biosecurity and governance ### Internal evaluations that offer partial reassurance **Anthropic** states that the models used by **Claude Science** underwent the company's standard safety and **biosecurity** evaluations before deployment in a sensitive **scientific** context, a particularly important precaution when dealing with tools capable of manipulating potentially dangerous genomic and chemical **data**. This transparency about internal procedures sets **Anthropic** apart from other technology players less inclined to publish their evaluation methodologies, a positive point in a sector where the race for speed could easily win out over caution. ### The ever-present risk of malicious dual use A tool capable of accelerating **research** in chemistry and biology necessarily carries a dual-use risk, where the same capabilities that enable the discovery of **treatments** could theoretically be diverted for malicious purposes, an issue the **Western** **scientific** community must take seriously starting now. **Anthropic** has not publicly detailed the precise monitoring **mechanisms** put in place to prevent such misuse, leaving an important governance question unresolved for the months ahead. *I believe vigilance on biosecurity must remain an absolute priority as these tools grow more powerful and more accessible. The West has a moral obligation to lead by example in the responsible governance of artificial intelligence applied to sensitive sciences.* ## The scientific community's mixed reception ### Cautious enthusiasm among early users Organizations that tested the platform during its beta phase reported significant efficiency gains, according to **Anthropic**'s own claims, though these testimonials come primarily from sources selected by the company itself rather than from independent evaluations conducted by neutral third parties. This methodological nuance does not invalidate the tool's potential benefits, but it calls for additional caution before accepting without reservation the performance figures put forward by the company in its own promotional communications. ### The questions the academic community keeps asking Several university **researchers** question the long-term sustainability of access to these tools, fearing that growing dependence on private commercial platforms could weaken the independence of public **scientific research** in the face of large technology companies' commercial interests. This tension between private innovation and public **scientific** independence is not new, but it takes on a particular dimension when the company in question is simultaneously preparing to become a competing pharmaceutical player itself. *I understand the academic community's reservations about this growing dependence on private platforms. But I also think refusing these tools on principle would deprive Western research of a real competitive advantage against rivals who will have no qualms whatsoever about adopting them en masse.* ## What this means for patients with neglected diseases ### Measured hope for often-forgotten populations **Neglected diseases** mainly affect populations in low- and middle-income countries, often abandoned by the traditional pharmaceutical industry due to a perceived lack of sufficient profit potential. If **Anthropic**'s program delivers on its promises, it could partially fill this glaring therapeutic void. It's worth remembering, however, that a stated intention guarantees no concrete outcome whatsoever, and that many similar initiatives announced by other technology companies in the past have not always delivered the promised **treatments** within the timelines originally suggested. ### The importance of measuring progress over several years Developing a drug, even with the most advanced artificial intelligence assistance, generally takes several years between the initial **discovery** of a promising candidate and its eventual regulatory approval, a timeline that far exceeds the current media cycle surrounding **Anthropic**'s IPO. Serious observers of the sector will therefore need to follow this story over several years rather than judging its success or failure in the months immediately following the initial announcement. *I believe the true measure of this initiative's success won't be visible for several years. Until then, I'd rather remain cautiously optimistic than give in to early enthusiasm that could prove premature given the real complexity of pharmaceutical discovery.* ## The broader context of the financialization of scientific research ### When science becomes a stock market argument This announcement fits into a broader trend in which major technology companies use their **scientific** capabilities as a sales pitch to investors, gradually turning fundamental **research** into a tool of financial communication rather than a purely academic end in itself. This growing financialization of science is not necessarily negative if it generates additional resources for **research**, but it deserves close monitoring to prevent **scientific** priorities from being dictated solely by short-term stock valuation considerations. ### A delicate balance to preserve for the future The challenge for **Anthropic**, as for the entire **Western** technology industry, is to maintain a delicate balance between the legitimate commercial imperatives of a private company and the **scientific** and ethical responsibilities that come with developing tools as powerful as those offered by **Claude Science**. This balance, if maintained with rigor, could serve as a positive model for the entire **Western** technology sector in the years ahead. *I believe this balance between commercial ambition and scientific responsibility will define Anthropic's long-term reputation far more than its current stock valuation. History rarely remembers companies that only made promises, but always remembers those that delivered verifiable results.* ## Conclusion: between scientific promise and financial maneuvering ### A bet worth watching closely The launch of **Claude Science** and the announcement of an internal drug development program represent a bold step for **Anthropic**, at a strategically crucial moment for the company ahead of its long-awaited IPO. The ambition to target **neglected diseases** deserves praise, but it must now be followed by concrete, verifiable **results** rather than mere statements of intent. For now, the company has delivered an impressive technical tool from a software engineering standpoint, but it has not yet demonstrated that it can turn this infrastructure into genuine therapeutic **discoveries** that will actually benefit patients. ### What the West should take away from this episode This announcement confirms that the race for **scientific** artificial intelligence is now a major strategic stake for **Western** technology companies, no longer just a marginal academic **research** topic. The West has every interest in encouraging this competition, provided it maintains rigorous standards of safety and scientific transparency. It remains to be seen whether **Anthropic** will manage to honor its humanitarian promises toward **neglected diseases**, or whether this initiative will remain, above all, a skillfully timed sales pitch ahead of one of the most anticipated IPOs of the current technology decade. *I'll close this commentary with a simple conviction: Anthropic's scientific ambition deserves to be praised with caution, not applauded without reservation. Patients with neglected diseases are not treated with press releases, but with verified treatments that, to this day, remain to be proven.* **By Maxime Marquette, columnist** ## Columnist's transparency note ### Editorial stance I am not a journalist but a **columnist and analyst**. My expertise lies in observing and analysing the geopolitical, economic and strategic dynamics that shape our world. My work is to take political strategies apart, to understand global economic movements, to put the decisions of international actors in context and to offer analytical perspectives on the changes that are redefining our societies. I do not claim the cold objectivity of traditional journalism, which stops at reporting facts. **I claim analytical clarity**, rigorous interpretation and a deep understanding of the complex issues that concern us all. My role is to give meaning to the facts, to place them in their historical and strategic context, and to offer a critical reading of events. ### Method and sources This text keeps the basic distinction between **verified facts and interpretive analysis**. The rule never changes: a factual claim is published only when a verifiable source supports it, and the sources this article actually used are listed under **Sources**, never here. **Primary source families** used by the publication, when the story has them: official statements from governments and international institutions, public statements by political leaders, reports from intergovernmental organisations, dispatches from recognised international news agencies. **Secondary source families**: specialised publications, internationally recognised news media, analyses from established research institutions, reports from industry organisations. When an article cites statistical, economic or geopolitical data, it comes from the institutions that produce it (intergovernmental organisations, central banks, national statistics offices), and the exact institution is named under **Sources**. ### Nature of the analysis The analyses, interpretations and perspectives in the analytical sections of this article are a **critical, contextual synthesis** based on the information available, the trends observed and the expert comments cited in the sources consulted. My role is to interpret these facts, to set them in the context of today's geopolitical and economic dynamics, and to give them a coherent meaning within the larger story of the changes shaping our time. These analyses reflect expertise built through continuous observation of international affairs and an understanding of the strategic mechanisms that drive global actors. This article describes a documented state of affairs on its publication date, not a prediction: later developments may shift its outlook. No update is promised in advance; when an article is corrected or completed, the change is dated in the text. ## Sources ### Primary sources - The Verge — Anthropic wants to develop its own drugs, July 3, 2026 - Reuters — Anthropic unveils Claude Science for scientific research, June 30, 2026 - TechCrunch — Claude Science bets on workflow rather than a new model, June 30, 2026 ### Secondary sources - Yahoo Finance — Anthropic launches Claude Science to diversify revenue ahead of its IPO - India Today — What to know about Anthropic's drug discovery program, July 1, 2026 - SmartAsset — Guide to Anthropic's IPO, valuation, and timeline - Pondero — Anthropic launches Claude Science, a workspace connected to more than 60 scientific databases --- Source: Maxime Marquette, MadMax. When citing, name the author and link https://mad-max.co/en/article/claude-science-le-pari-risque-d-anthropic-avant-son-entree-en-bourse