Anthropic PBC has unveiled Claude Science, a new artificial intelligence platform intended to streamline the research process for scientists by automating repetitive and labour-intensive components of their work. The tool, which became available to paying users on June 30, consolidates access to more than 60 scientific databases and permits researchers to pose queries in plain language rather than navigating multiple disparate information sources individually. This integration aims to simplify multistep research workflows that would otherwise demand considerable manual effort and cross-referencing across various platforms.
The capabilities built into Claude Science extend to substantial technical tasks within the life sciences domain. Scientists can leverage the platform to predict protein structures and conduct complex analyses spanning both biology and chemistry disciplines. By automating these traditionally time-consuming processes, Anthropic contends that researchers can redirect their efforts toward higher-level analytical and conceptual work rather than data gathering and preliminary processing. The company emphasised that Claude Science generates outputs accompanied by traceable documentation, enabling researchers to verify the accuracy and validity of the information produced by the system.
Anthropically's announcement of Claude Science coincided with a broader strategic commitment to drug discovery development. The organisation revealed plans to establish internal preclinical drug discovery programmes targeting therapeutic areas that conventional pharmaceutical and biotechnology companies have historically deemed unattractive or commercially unviable. This represents a significant expansion beyond software provision into active biopharmaceutical research and development, positioning Anthropic as a direct participant in the discovery pipeline rather than merely an infrastructure provider.
The San Francisco event where Claude Science was announced gathered influential figures from both the technology and pharmaceutical sectors. Anthropic's chief executive officer Dario Amodei shared the stage with Vas Narasimhan, who leads Novartis AG and serves on Anthropic's board, alongside Chris Boerner, the chief executive of Bristol-Myers Squibb Co. This assembly of executives underscores the convergence of interests between artificial intelligence innovators and established pharmaceutical manufacturers seeking to harness computational tools for accelerated discovery timelines.
Anthropic's valuation has reached US$965 billion, positioning it among the world's most valuable artificial intelligence enterprises. The company is pursuing an initial public offering potentially as early as autumn, making the demonstration of tangible commercial applications increasingly critical for justifying its market capitalisation to investors. Within this context, Claude Science serves as evidence that Anthropic's technology produces practical utility across professional sectors, moving beyond experimental prototypes toward products with immediate business applications.
The competitive landscape between Anthropic and rival OpenAI has intensified considerably over the past eighteen months as both firms race to develop specialised artificial intelligence tools tailored to diverse professional domains. These efforts span financial services, legal work, healthcare, and scientific research—sectors characterised by high-value transactions and significant demand for knowledge work automation. Both companies are fundamentally attempting to justify premium valuations by demonstrating that their platforms generate substantial economic value and can capture meaningful market share across multiple industries simultaneously.
However, recent product launches from Anthropic have triggered pronounced market concerns about technological disruption and employment displacement. When the company introduced Claude Cowork in February, designed to automate legal functions such as contract review and legal research, global stock markets experienced a notable contraction. The announcement apparently crystallised broader investor anxieties regarding which professional service categories might face obsolescence as artificial intelligence capabilities mature. This US$1 trillion market rout illustrated how announcements of automation tools can rapidly reshape market valuations across multiple sectors, reflecting fundamental uncertainty about future labour demand in knowledge-intensive professions.
Dario Amodei acknowledged the necessity of translating technological capability into tangible clinical outcomes. He articulated an expectation that within the coming year, Anthropic would demonstrate meaningful progress in utilising artificial intelligence to identify novel drug targets, though he stopped short of committing to specific outcomes. This cautious framing suggests awareness that pharmaceutical innovation involves substantial uncertainty and lengthy development timelines, constraining the pace at which artificial intelligence applications can demonstrably influence patient outcomes.
Vas Narasimhan struck a more measured tone regarding artificial intelligence development and deployment, emphasising that the pharmaceutical and biotechnology sectors must deliver concrete benefits to patients rather than merely issuing aspirational statements about artificial intelligence's potential. He further advocated for proactive regulatory frameworks governing artificial intelligence deployment, warning that absent appropriate governance mechanisms, a crisis might ultimately force governments to implement restrictive regulations. His remarks reflect pharmaceutical sector leadership's recognition that artificial intelligence tools require careful oversight to maintain public trust and ensure responsible development.
Claude Science incorporates Anthropic's existing Claude models, specifically Opus 4.8 which the company released in May. The platform's outputs include detailed documentation permitting scientists to trace how the artificial intelligence arrived at particular conclusions, while images generated through the system contain embedded information about the computational processes underlying their creation. This emphasis on transparency and traceability addresses scientific communities' legitimate concerns about verifiability and reproducibility in artificial intelligence-assisted research.
The timing of Claude Science's release carries geopolitical significance. Merely thirteen days prior, Anthropic had restricted access to its most advanced models, Fable 5 and Mythos 5, in compliance with Trump administration directives aimed at preventing foreign nationals from utilising cutting-edge American artificial intelligence technology. Subsequently, the company obtained United States government approval to restore partial access to Mythos 5 after resolving national security concerns. These regulatory manoeuvres underscore how artificial intelligence development now intersects directly with international relations and security policy, constraining how and where advanced models can be deployed. For Malaysia and Southeast Asian nations, such restrictions carry implications regarding technology access and the extent to which regional companies and researchers can participate in artificial intelligence-driven innovation without facing geopolitical barriers.
