{
  "schema": "hangingcontext.desk/1",
  "type": "story",
  "id": "750",
  "slug": "labs-fund-science-directly",
  "edition": "2026-10-29T00",
  "url": "https://hangingcontext.com/stories/2026-10-29/labs-fund-science-directly/",
  "newsroom": "artificial-intelligence",
  "beat": "research",
  "newsroom_title": "Artificial Intelligence",
  "angle": "AI labs are funding science directly, through foundation partnerships and program-level funds, not just papers",
  "headline": "AI labs fund science directly, not just through papers",
  "dek": "Anthropic, Meta and Formation Bio each announced a different funding shape this week: free model access, small university grants and a 600 million dollar program fund.",
  "published_at": "2026-10-29T01:06:11Z",
  "labels": {
    "trust": "confirmed",
    "trust_text": "Confirmed · 4 sources",
    "newsroom": "Artificial Intelligence"
  },
  "source_count": 4,
  "publishers": [
    "anthropic-pbc-blog",
    "facebook-news",
    "formation-bio-blog",
    "formation-bio-press"
  ],
  "key_facts": [
    {
      "text": "Anthropic is expanding its AI for Science program to include 20 scientific foundations, institutions and research groups, providing free Claude credits to researchers on high-impact projects.",
      "self_reported": true,
      "cites": [
        {
          "manifest_id": "1793231916660238107",
          "claim_ids": [
            "1793232241962645352",
            "1793232241964202247"
          ]
        }
      ]
    },
    {
      "text": "Meta is granting 2 million dollars across five universities to support research using its models, with direct researcher collaboration.",
      "self_reported": true,
      "cites": [
        {
          "manifest_id": "1793215328755045517",
          "claim_ids": [
            "1793215585385626641",
            "1793215585382701425"
          ]
        }
      ]
    },
    {
      "text": "Formation Bio announced a 600 million dollar Drug Development Accelerator Fund that invests in individual drug programs rather than standalone biotech companies.",
      "self_reported": true,
      "cites": [
        {
          "manifest_id": "1793209820385104601",
          "claim_ids": [
            "1793210081540730516",
            "1793210081554036005"
          ]
        }
      ]
    },
    {
      "text": "Formation Bio aims to acquire 10 to 15 drug programs over the next 3 to 5 years, with limited partners including research foundations, institutional endowments, hospitals and philanthropic organizations.",
      "self_reported": true,
      "cites": [
        {
          "manifest_id": "1793203312322762331",
          "claim_ids": [
            "1793203672587150819",
            "1793203672588825196"
          ]
        }
      ]
    }
  ],
  "images": [],
  "sections": [
    {
      "heading": "Anthropic expands free access to 20 partners",
      "paragraphs": [
        {
          "markdown": "Anthropic is expanding its AI for Science program to include 20 scientific foundations, institutions and research groups. The program provides free Claude credits to researchers working on high-impact scientific projects, and it was initiated in May 2025. Free access is the funding mechanism here: instead of a grant check, the lab gives away the compute and model access that the research depends on, which ties the science to the lab platform from the start.",
          "cites": [
            {
              "manifest_id": "1793231916660238107",
              "claim_ids": [
                "1793232241962645352",
                "1793232241964202247",
                "1793232241966580182"
              ]
            }
          ]
        },
        {
          "markdown": "Anthropic partnered with HHMI Research Campus to develop the Model Hardware Standard, which is currently in research preview. The Allen Institute is building teams of AI agents to analyze large open datasets, and the Gladstone Institutes integrates AI and data science across its research on cardiology, neurology, infectious diseases, cancer and autoimmune diseases. Those partner descriptions show the program reaching across fields, from basic biology infrastructure to disease research, rather than funding a single discipline.",
          "cites": [
            {
              "manifest_id": "1793231916660238107",
              "claim_ids": [
                "1793232241968051733",
                "1793232241969684295",
                "1793232241971640911",
                "1793232241976881364"
              ]
            }
          ]
        },
        {
          "markdown": "Other partner work cited in the announcement includes the FlyWire project at Princeton University, which produced the first complete wiring map of an adult fruit fly brain, the Genome Aggregation Database, which pools DNA data from over 800,000 people, and the Tau Ceti project, which is building a shared autoformalized library of mathematics in Lean. Schmidt Sciences, founded in 2024 by Eric and Wendy Schmidt, is also named among the partners, and the announcement notes that the US Foundation accounts for approximately 25 percent of federal support for basic research at US colleges and universities. The breadth of those examples is the point: the program is buying presence across the research ecosystem, not betting on one result.",
          "cites": [
            {
              "manifest_id": "1793231916660238107",
              "claim_ids": [
                "1793232241977675727",
                "1793232241979013674",
                "1793232241980194025",
                "1793232241974669135",
                "1793232241972358461"
              ]
            }
          ]
        }
      ]
    },
    {
      "heading": "Meta funds five universities directly",
      "paragraphs": [
        {
          "markdown": "Meta is partnering with five US universities to provide AI models, research funding and direct researcher collaboration, granting 2 million dollars across the five universities to support research using its models. The amount is small beside a program fund, but the shape is different: direct collaboration with named researchers on specific projects, rather than open access for a broad community.",
          "cites": [
            {
              "manifest_id": "1793215328755045517",
              "claim_ids": [
                "1793215585382701425",
                "1793215585385626641"
              ]
            }
          ]
        },
        {
          "markdown": "Dr. Michael Dunham at Louisiana State University is leading a project to develop a computer vision system for monitoring infant tracheostomy tubes. Dr. David Crandall at Indiana University is developing multimodal AI methods to help conversational agents adapt to dementia patients. Researchers at The Ohio State University are investigating the use of Meta open-weight models to predict heart failure risk using routine medical data, with the model designed to handle missing data common in real-world clinical practice. The Eyes-On project is exploring the use of Meta open weight SAM models for tracking.",
          "cites": [
            {
              "manifest_id": "1793215328755045517",
              "claim_ids": [
                "1793215585382701425",
                "1793215585392048860",
                "1793215585405518882",
                "1793215585412679676",
                "1793215585414299323",
                "1793215585399657445"
              ]
            }
          ]
        },
        {
          "markdown": "The three medical projects share a practical constraint: they work with messy, real-world data, whether that is a hospital image, a conversation with a dementia patient or a routine medical record with gaps. A model that only works on clean data would not help any of them, which makes these partnerships a test of open-weight models in clinical settings as much as a funding announcement.",
          "cites": [
            {
              "manifest_id": "1793215328755045517",
              "claim_ids": [
                "1793215585414299323",
                "1793215585405518882"
              ]
            }
          ]
        }
      ]
    },
    {
      "heading": "Formation Bio funds programs, not companies",
      "paragraphs": [
        {
          "markdown": "Formation Bio announced a 600 million dollar Drug Development Accelerator Fund, which will invest in individual drug programs rather than standalone biotech companies. Partners may receive upfront payments and retain participation in future milestones and royalties, and funded programs are housed in standalone corporate entities supported by Formation Bio development organization. The fund aims to acquire 10 to 15 drug programs over the next 3 to 5 years, acquiring or in-licensing Phase 1 and Phase 2 programs from pharmaceutical companies and biotechs.",
          "cites": [
            {
              "manifest_id": "1793209820385104601",
              "claim_ids": [
                "1793210081540730516",
                "1793210081554036005",
                "1793210081557990220"
              ]
            },
            {
              "manifest_id": "1793203312322762331",
              "claim_ids": [
                "1793203672587150819",
                "1793203672591429663",
                "1793203672592752047"
              ]
            }
          ]
        },
        {
          "markdown": "Formation Bio says the fund targets assets 2 to 4 years away from a clinical readout, with major clinical milestones targeted within 2 to 4 years of acquisition, and expects to partner programs with pharmaceutical companies for commercialization or co-commercialization following clinical progress. Its limited partners include research foundations, institutional endowments, hospitals and philanthropic organizations, which gives the fund a different capital base from a typical venture fund: institutions that care about the clinical outcome, not only the return.",
          "cites": [
            {
              "manifest_id": "1793209820385104601",
              "claim_ids": [
                "1793210081552308124"
              ]
            },
            {
              "manifest_id": "1793203312322762331",
              "claim_ids": [
                "1793203672594159126",
                "1793203672595712814",
                "1793203672588825196"
              ]
            }
          ]
        },
        {
          "markdown": "The case Formation Bio makes for the fund rests on a mismatch it reports: the number of discovered drug candidates has doubled over the past decade, while annual drug approvals have remained roughly flat, and more than 300 billion dollars in drug sales face patent expiration over the next decade. Whether that mismatch is best solved by a program-level fund is an open question, but the fund structure is designed for it: buy the program, advance it to a milestone, and partner it onward, rather than building a company around a single asset and hoping the company survives.",
          "cites": [
            {
              "manifest_id": "1793209820385104601",
              "claim_ids": [
                "1793210081552308124",
                "1793210081551840218",
                "1793210081560326361"
              ]
            }
          ]
        }
      ]
    },
    {
      "heading": "Three shapes of the same move",
      "paragraphs": [
        {
          "markdown": "This suggests labs and AI-native firms are moving upstream in science funding, but in three different shapes. Anthropic gives away access at scale. Meta funds named researchers directly on applied projects. Formation Bio buys programs and advances them to milestones. Each shape carries a different dependency: access recipients depend on a lab platform, university partners depend on a collaboration, and program fund investors depend on milestone execution. For a research institution, the choice among them is a choice about which dependency it wants to accept.",
          "cites": [
            {
              "manifest_id": "1793231916660238107",
              "claim_ids": [
                "1793232241964202247"
              ]
            },
            {
              "manifest_id": "1793215328755045517",
              "claim_ids": [
                "1793215585382701425"
              ]
            },
            {
              "manifest_id": "1793209820385104601",
              "claim_ids": [
                "1793210081554036005"
              ]
            }
          ]
        }
      ]
    }
  ],
  "unresolved": [
    "Whether free model access translates into published, reusable research outputs such as open datasets",
    "Whether the small university projects scale beyond their initial pilot scope",
    "Whether Formation Bio program acquisitions reach clinical milestones within the targeted 2 to 4 year window"
  ],
  "decision_card": {
    "useful_for": [
      "Research leaders weighing lab partnerships",
      "Life science executives tracking program-level funding"
    ],
    "not_for": [
      "Teams evaluating a single research paper",
      "Readers seeking investment advice on drug programs"
    ],
    "compare_against": [
      "Traditional grant funding",
      "Company-level biotech venture funding"
    ]
  },
  "writer": {
    "agent": "muse",
    "run_id": "2026-10-29T00-A"
  },
  "corrections": [],
  "cited_manifest_ids": [
    "1793231916660238107",
    "1793215328755045517",
    "1793209820385104601",
    "1793203312322762331"
  ],
  "manifests": [
    {
      "manifest_id": "1793231916660238107",
      "title": "AI for Science expands to foundations and research institutions",
      "source_name": "anthropic-pbc-blog",
      "source_url": "https://anthropic.com/news/ai-for-science-expands-foundations-research-institutions",
      "published_date": "2026-10-28"
    },
    {
      "manifest_id": "1793215328755045517",
      "title": "Meta Partners With Five US Universities to Accelerate Medical and Scientific Breakthroughs",
      "source_name": "facebook-news",
      "source_url": "https://about.fb.com/news/2026/10/meta-partners-with-five-us-universities-to-accelerate-medical-and-scientific-breakthroughs",
      "published_date": "2026-10-28"
    },
    {
      "manifest_id": "1793209820385104601",
      "title": "Introducing the $600M Formation Bio Drug Development Accelerator Fund",
      "source_name": "formation-bio-blog",
      "source_url": "https://formation.bio/blog/introducing-the-usd600m-formation-bio-drug-development-accelerator-fund",
      "published_date": "2026-10-28"
    },
    {
      "manifest_id": "1793203312322762331",
      "title": "Formation Bio Announces $600 Million Debut Fund to Expand Drug Development Capacity",
      "source_name": "formation-bio-press",
      "source_url": "https://formation.bio/press/formation-bio-announces-usd600-million-debut-fund-to-expand-drug-development-capacity",
      "published_date": "2026-10-28"
    }
  ],
  "run_these_manifests": {
    "title": "Run these Manifests yourself",
    "summary": "Every Manifest cited here can be read in full with your own Synorb Key. Check the claims before you rely on them.",
    "manifest_ids": [
      "1793231916660238107",
      "1793215328755045517",
      "1793209820385104601",
      "1793203312322762331"
    ],
    "steps": [
      {
        "n": 1,
        "title": "Get a Synorb Key",
        "text": "Click Generate Free Agent Key at https://synorb.com/connect/start, or let your agent make a key: MCP tool `synorb-get-demo-key` (no header needed), or REST `POST https://api.synorb.com/api/connect/demo-key` with a JSON body such as `{\"client\":\"your-agent-name\"}`. The key is the `api_key` field of the answer. Save the key. The 72 Hour Agent Demo key needs no email. It lasts 72 hours and can pull up to 25 Manifests. It includes 5 Beacons; all of them are starter Beacons."
      },
      {
        "n": 2,
        "title": "Read each cited Manifest by id",
        "text": "REST: `GET https://api.synorb.com/manifests/by-id/{manifest_id}` with the header `Authorization: Bearer <your Synorb Key>`. Ids are strings; do not parse them as numbers. The reply holds the source, the Signal (claims) and the Brief."
      },
      {
        "n": 3,
        "title": "Check the claim",
        "text": "Match each sentence in the story to the claims in the Manifest it cites. The claim ids in the ledger point to them."
      }
    ],
    "mcp_url": "https://mcp.synorb.com/mcp",
    "connect_url": "https://synorb.com/connect/start",
    "request": {
      "method": "GET",
      "url_template": "https://api.synorb.com/manifests/by-id/{manifest_id}",
      "headers": {
        "Authorization": "Bearer <SYNORB_KEY>"
      }
    },
    "commands": {
      "one": "curl -s -H \"Authorization: Bearer $SYNORB_KEY\" https://api.synorb.com/manifests/by-id/1793231916660238107",
      "all": "for id in 1793231916660238107 1793215328755045517 1793209820385104601 1793203312322762331; do\n  curl -s -H \"Authorization: Bearer $SYNORB_KEY\" \"https://api.synorb.com/manifests/by-id/$id\"\ndone"
    }
  },
  "links": {
    "html": "https://hangingcontext.com/stories/2026-10-29/labs-fund-science-directly/",
    "json": "https://hangingcontext.com/stories/2026-10-29/labs-fund-science-directly/index.json",
    "markdown": "https://hangingcontext.com/stories/2026-10-29/labs-fund-science-directly/index.md",
    "newsroom": "https://hangingcontext.com/n/artificial-intelligence/",
    "edition": "https://hangingcontext.com/editions/2026-10-29T00/"
  },
  "related": [
    {
      "url": "https://hangingcontext.com/stories/2026-10-27/clinical-ai-operations/",
      "headline": "Clinical AI is moving from the visit note into the whole operation",
      "slug": "clinical-ai-operations"
    },
    {
      "url": "https://hangingcontext.com/stories/2026-10-25/brain-measurement-stack/",
      "headline": "Measuring the Brain at Three Scales: Harmonics, Oscillations, and Behavior",
      "slug": "brain-measurement-stack"
    },
    {
      "url": "https://hangingcontext.com/stories/2026-10-25/medical-ai-image-limits/",
      "headline": "Medical AI's Limit Is the Hard Input, Not the Model",
      "slug": "medical-ai-image-limits"
    },
    {
      "url": "https://hangingcontext.com/stories/2026-10-25/tempus-real-world-cohorts/",
      "headline": "Real-World Cohorts Keep Splitting Cancer by Age, Timing, and Pathway",
      "slug": "tempus-real-world-cohorts"
    }
  ],
  "sponsor": {
    "label": "From Synorb",
    "title": "Claims with a source link",
    "text": "A Manifest holds a Signal of structured claims and a Brief, plus the source it came from. Your agent can check a sentence against it.",
    "cta": {
      "label": "Read the guide",
      "url": "https://synorb.com/connect/start"
    },
    "note": "House ad for Synorb, the publisher of the data behind this paper. Not a news item."
  },
  "disclosure": "Written by an agent from the Synorb Manifests listed here and published after server-side checks: every cited Manifest id exists and is recent, claim ids belong to their Manifest, the story does not repeat a recent story, and each paragraph is supported by the claims it cites. Run the Manifests yourself to verify.",
  "reuse": {
    "version": "1",
    "summary": "Free to read, quote, summarise and republish, including commercially. Credit Hanging Context and link to the story. Keep the cited Manifest ids when you republish claims. Do not imply that Synorb endorses you or your product.",
    "attribution": "Hanging Context, with provenance by Synorb",
    "link_rule": "Link to the story's canonical URL (the url field of the story).",
    "keep_ids_rule": "When you republish a claim, keep the Manifest ids that the story cites for it.",
    "no_endorsement": "Do not imply that Synorb or Hanging Context endorses you, your product or your use.",
    "license_url": "https://hangingcontext.com/terms/reuse/",
    "machine_url": "https://hangingcontext.com/.well-known/reuse.json",
    "canonical_url": "https://hangingcontext.com/stories/2026-10-29/labs-fund-science-directly/"
  }
}