AI

Formation Bio raises $372M to boost drug development with AI

Formation Bio, a startup focused on applying AI to drug development with backing from OpenAI CEO Sam Altman, has raised over a quarter-billion dollars to support its ambitious product roadmap.

Formation announced Wednesday that it raised $372 million in a Series D funding round led by Andreessen Horowitz with participation from drug maker Sanofi, Sequoia, Thrive, Emerson Collective, Lachy Groom, SV Angel Growth and FPV Ventures. The new tranche brings Formationโ€™s total raised to more than $600 million (according to PitchBook), which the company says is being put mainly toward partnership acquisition efforts and R&D.

Formation declined to reveal its new valuation. But a spokesperson told TechCrunch that itโ€™s a โ€œmaterial step upโ€ from $1 billion, Formationโ€™s Series C valuation.

The company, which previously went by the brand TrialSpark, was co-founded by Benjamine Liu and Linhao Zhang in 2016. Liu has a background in computational biology, having conducted neuroscience research at Oxford and UPenn. Zhang is a software developer by trade and worked at Salesforce before joining Oscar Insurance as a product engineer.

Formation builds tech-forward solutions for clinical trials and drug development. The company licenses drug IP from and co-develops drugs with biotech and pharma companies, and develops these drugs past clinical proof-of-concept.

Drug development is a notoriously expensive and challenging endeavor. It takes 10 to 15 years on average to take a drug from initial discovery through regulatory approval, with the cost per drug reaching up to $5.5 billion. And an estimated 90% of drugs fail to reach the market. 

Formation claims itโ€™s able to run clinical trials more efficiently by streamlining processes such as study startup, participant recruitment and data management. For example, the company is currently deploying AI to generate patient recruitment materials and reports for โ€œadverse events.โ€ Itโ€™s also fine-tuning AI models to provide drug development teams recommendations for R&D decisions and better predict drug toxicity, tolerability and efficacy.

0 of 38 minutes, 24 secondsVolume 0%
00:04
38:20
38:24
 

Last month, Formation announced a collaboration with OpenAI and Sanofi to jointly design and develop customized AI solutions for drug development. OpenAI said it would contribute access to AI capabilities and expertise, and Sanofi said it would bring proprietary data for developing AI models.

OpenAIโ€™s involvement gives the appearance of conflict of interest, given that Altman was involved in Formationโ€™s Series C fundraising. OpenAI PR told TechCrunch the deal was led by OpenAI COO Brad Lightcap and OpenAIโ€™s board of directors but didnโ€™t indicate whether Altman, whoโ€™s on the board, recused himself.

โ€œAt Sanofi, weโ€™re all in on AI,โ€ Sanofi CEO Paul Hudson said in a press release. โ€œAnd we are proud to partner with and invest in Formation Bio, whose AI-driven drug development vision and capabilities will help lead our industry forward in the shared ambition to accelerate and improve how we bring more new medicines to patients.โ€

Formation has three drug candidates in its clinical pipeline, including treatments for chronic hand eczema, sensory neuropathy and knee osteoarthritis. The furthest along is the eczema treatment, which recently reached phase 3 โ€” the last stage of testing before a drug is submitted to regulatory authorities.

A number of startups are attempting to pioneer AI-powered tech for drug development, including EvolutionaryScale, which emerged from stealth this week with investments from Amazon and Nvidia. Market research firm Markets and Markets anticipates that the market for AI in drug discovery will be worth $4.9 billion by 2028. Major players in the space include Xaira (which launched with $1 billion), DeepMind spin-off Isomorphic, Insilico, Jeff Dean-backed Profluent, Enveda and Causaly.

Topics

, , , , , , , , , ,
Keep reading
Image Credits:Stilta
Venture

Stilta raises $10.5M from a16z and YC to help companies rediscover the patents they forgot they had

Oskar Block has never been able to stay away from entrepreneurship for long.

He was just 18 when he launched his first startup, building machine learning models for sports betting. โ€œIโ€™ve always been drawn to solving difficult data problems,โ€ he told TechCrunch. He went on to consulting, where he helped companies with their AI integration strategies and learned what it took to get large enterprises to embrace the technology. 

Block then took a role at an autonomous trucking company, where he saw firsthand how manual and slow the patent process was. The idea for his next company came one evening at dinner with a friend and colleague, Tobias Estreen, when Estreenโ€™s father, a patent attorney, started recounting what his days looked like: โ€œReading the same kind of documents, the same way he had for thirty years,โ€ Block recalled.

Block and Estreen saw an opening and teamed up with two others, Petrus Werner and Oscar Adamsson, to launch Stilta, an AI platform designed to automate the research and analytical work behind intellectual property cases โ€” the kind of labor-intensive work that has historically made patent litigation slow and expensive. The startup announced a $10.5 million seed round on Tuesday, led by Andreessen Horowitz. Other investors include Y Combinator and operators from companies like OpenAI, Legora, and Lovable. 

Block, the companyโ€™s CEO, said Stilta works like a team of lawyers. Users put a patent number into the software along with any relevant content, and from there, a network of AI agents gets to work, searching for other patents might conflict with the claim, flagging similar property that could apply, and pulling the filing and court history of the patent. 

โ€œThey reason in parallel and converge the way a room full of specialists would, but at a scale no human team can match,โ€ Block said, adding that the lawyer or professional using the platform is still in the โ€œdriverโ€™s seatโ€ by guiding the analysis, not ceding it. โ€œThe output is litigation-grade: a report and claim charts with pinpoint citations to every piece of evidence.โ€

Other companies in this space include Solve Intelligence and DeepIP. Legal tech has become a hot sector amid the broader AI boom. Block said parts of the legal industry are already seeing AI-accelerated change, while other parts may not be ready for it for a long time. 

The Musk v. Altman case is in the jury's hands. The trust verdict is still out. | Equity Podcast
0 seconds of 38 minutes, 24 secondsVolume 0%
 

The analytical work, he said, is already being overtaken by AI. For now, itโ€™s still humans deciding the outcomes of cases. He also noted that many companies are holding on to patents theyโ€™ve โ€œnever enforced, never licensed, never even analyzed properly because the cost of doing so was prohibitive.โ€

That cost barrier is what Stilta aims to lower. Making the patent litigation process more efficient and affordable could open up new doors for many companies that have long left their IP on the shelf and change how they think about the latent value sitting inside their patent portfolios.

โ€œThe question isnโ€™t really whether the legal system is prepared for AI,โ€ Block said. โ€œItโ€™s whether companies are prepared for what becomes possible when the analytical bottleneck disappears.โ€ 

Topics

, , , ,
Keep reading
Image Credits:Google/screenshot /
AI

OpenAI is making it easier to check if an image was made by their models

With AI image generators widely available online and more sophisticated than ever, itโ€™s never been harder to tell if an image is authentic. But on Tuesday, OpenAI announced two new measures to help fight the problem.

The company has committed to an open standard called C2PA, which adds a clear signal in metadata that an image was generated by AI. OpenAI is also partnering with Google to include an invisible watermark called SynthID, which will be harder to detect, but also harder to erase if bad actors try to cover their tracks.

The new protections only apply to images generated by OpenAI products, so they wonโ€™t affect the flood of imagery coming from less reputable AI tools; they can help ensure that OpenAI isnโ€™t part of the problem.

OpenAI is also previewing a public verification tool that will check for both signals, allowing users to easily test whether an image was generated using AI. Initially, the tool will only extend to images generated by OpenAI products; the company hopes to expand it to cover other tools over time.

Founded in 2021, the Coalition for Content Provenance and Authenticity (C2PA) is a non-profit dedicated to mitigating the harmful effects of AI imagery on public discourse. The C2PA standard has been adopted by a range of Google products, but adoption remains inconsistent across the industry. Because the C2PA signal is clearly accessible in the metadata of each file, it can be manipulated, and is most useful among trusted users.

SynthID is a newer effort designed to be a more robust measure to meddling. Developed by Google, the SynthID watermark is designed to persist even when bad actors attempt to remove it, either through screenshots, resizing or digital manipulation.

The two systems are meant to complement each other, with each addressing the otherโ€™s weaknesses.

The Musk v. Altman case is in the jury's hands. The trust verdict is still out. | Equity Podcast
0 seconds of 38 minutes, 24 secondsVolume 0%
 

โ€œWatermarking can be more durable through transformations like screenshots, while metadata can provide more information than a watermark alone,โ€ OpenAI noted in its announcement. โ€œTogether, they make provenance more resilient than either layer would be on its own.โ€

Topics

, , , , , ,
Loading the next article
Some areas of this page may shift around if you resize the browser window. Be sure to check heading and document order.