BIO Token Rises 105% as AI-Driven Drug Discovery Attracts Attention

Author: Muhammad Yusuf,Research @Delphi_Digital

Compiled by: Deep潮 TechFlow

DeepChain Summary: The AI agent peptAI designed a candidate drug for ADHD in 24 hours, causing the BIO Protocol token to surge 105%, and crypto Twitter was instantly flooded with "DeSci" posts. But stepping back to examine the data: while AI-discovered drugs show an 80–90% success rate in Phase I trials, these trials only test whether a drug is lethal—not whether it actually treats the disease. Of the 173 AI-driven drug pipelines globally, none have completed Phase III trials, which determine real therapeutic efficacy. The most optimistic projections suggest the first FDA approval won’t come until 2028, yet the crypto market’s attention span rarely lasts six months. The ultimate outcome of this game hinges on one of two possibilities: either DeSci develops a funding mechanism aligned with the four-year clinical trial cycle, or this becomes another narrative bubble that bursts.

An AI agent called peptAI designed a novel peptide candidate drug for ADHD from scratch in 24 hours, completed eight validation steps, and produced a molecule ready for wet-lab testing. The lab work cost only a few thousand dollars. BIO Protocol’s token surged 105%, and within hours, half of crypto Twitter’s bios added "DeSci," just as they had added "AI" six months ago.

Open-source protein folding models now achieve AlphaFold3-level performance at zero licensing cost, publicly available bioactive databases cover 2.5 million compounds, and wet-lab validation costs less than $2,000. AI is compressing the cost and time of drug discovery.

I spent a week trying to figure out what was truly different this time.

The commonly cited figure is that AI-discovered drugs have an 80–90% success rate in Phase I clinical trials, compared to about 47% for traditional drugs. But what’s often left unsaid is that Phase I only tests whether a drug will kill you, not whether it can cure you. Passing Phase I simply means your compound is safe enough to proceed with further research—but you still have to complete all remaining steps until FDA approval.

Fewer than 40 AI-discovered compounds have reached Phase II trials, and none have completed Phase III. Rentosertib from Insilico Medicine is currently the most advanced AI-discovered compound, with positive Phase IIa results for idiopathic pulmonary fibrosis published in Nature Medicine in mid-2025, and Phase III enrollment set to begin in China in Q4 2025. If everything proceeds perfectly (enrollment completed by 2027, data announced in 2028, FDA review in 2029), you’re looking at at least three more years—and this is still the best candidate among 173 drug pipelines, several of which were put on hold in 2025 after failing endpoints in atopic dermatitis, schizophrenia, and cancer. Independent analysts estimate a 60% probability that the first AI-designed drug will receive FDA approval by 2027; so far, no AI-designed drug has completed this process.

Keep these timelines in mind, then take a look at BIO Protocol’s candlestick chart. The token dropped from $0.89 to $0.018, then surged 105% after the peptAI announcement, with $720 million in trading volume swirling within a $68 million market cap. The entire DeSci funding logic assumes token holders will patiently wait 7 to 10 years for clinical trials, while Crypto Twitter moves on to the next narrative long before Phase I data is unblinded.

Pump Science leaked its private key on GitHub, spawning a slew of scam tokens, including one actually called Cocaine. The enforceability of IP-NFTs has never been tested in court.

If we avoid reflexive self-deception and chronic speculation, DeSci holds a glimmer of hope in open-source science.

In October 2025, the OpenFold Consortium released OpenFold3 under the Apache 2.0 license—fully trainable, commercially usable, and built on over 300,000 experimentally determined structures (unlike AlphaFold3, which Google locked behind academic-only restrictions). Boltz-2, jointly launched by MIT and Recursion, predicts protein structures and binding affinities 1,000 times faster than physics-based methods. The Baker Lab released RFdiffusion3 in December. ChEMBL offers 2.5 million bioactive compounds with complete ADMET profiles, freely available to anyone with a laptop. Infrastructure that once cost pharmaceutical companies millions to build now sits on GitHub under permissive licenses, and five pharmaceutical companies are currently conducting federated training on their proprietary drug-protein libraries through the federal OpenFold3 initiative. No one on Crypto Twitter is talking about this because there are no tokens to trade, and I genuinely doubt the core contributors to these codebases would be excited about tokens.

In 2025, over 7,800 NIH and NSF grants were terminated or paused, with more than $5 billion in funding frozen. The NIH (National Institutes of Health, the world’s largest public funder of biomedical research, with an annual budget of approximately $47 billion) saw its budget held flat as Congress continued appropriations, but the government still froze pipeline funding. New competitive grants plummeted from 11,659 in fiscal year 2024 to 6,095 in fiscal year 2025, a 48% decline. The success rate for researchers applying for funding dropped from 21% to 13%. Fred Hutch lost $508 million, and Harvard lost $945 million.

The funding vacuum is why DeSci has an opportunity to make an impact—if done right. In July 2025, VitaDAO-funded Gero signed a research and licensing agreement with Chugai Pharmaceutical (a Roche subsidiary with a market cap of ~$100 billion, not a meme), with milestone payments reaching up to $250 million. This was the first time a DAO-funded project produced something a real pharmaceutical company was willing to pay nine figures for. The process was completed without governance infighting or rug pulls, and it remains the most significant event to have occurred in this space.

This year, 15 to 20 AI-discovered drugs are entering Phase III; Rentosertib’s data won’t be available until at least 2028, meaning it will still be years before we truly know whether any of this translates into effective human therapies. Whether or not tokens exist, open-source tech stacks will continue driving down costs, and the funding vacuum will keep pushing researchers toward anyone willing to write a check. Open-source protein folding models now match AlphaFold3 performance at zero licensing cost, with wet-lab validation under $2,000, while the NIH has just recorded its lowest grant success rate in two decades. Even if AI delivers everything its proponents promise—halving the drug development timeline—you’re still looking at a 4- to 5-year journey from discovery to approval, and that’s under the optimistic assumption that Phase III success rates actually improve. Four years is an eternity in an industry where portfolio beliefs shift with quarterly earnings calls, and token holders treat a six-month holding period as a life sentence.

The cost of discovery and innovation is becoming cheaper each quarter, regardless of whether anyone is buying tokens. The funding vacuum caused by NIH is also an issue. Between these two factors, there may be a viable model where tokens fund specific clinical trials, set clear milestones, and delegate governance to experts. The Gero/Chugai deal is the first proof that DAO-funded projects can produce something a real pharmaceutical company is willing to pay nine figures for. Beyond the hype, I’m curious whether anyone will build defensible financing infrastructure for genuine DeSci.