LivingMind AI

LivingMind AI is an academic initiative founded by Dr. Benoit Coulombe in 2025 (see biography below).

LivingMind is a new sort of AI that is aligned with life. An AI that is rooted in biological principles, co-evolving with society and science, governed ethically, transparently and collectively, and designed to strengthen life - not dominate it. In LivingMind AI, control mechanisms ensuring security are integrated from the beginning and during AI development itself, implementing "governance-by-design".

SHORT BIOGRAPHY

Benoit Coulombe, PhD, is a molecular biologist and biochemist at the Montreal Clinical Research Institute (IRCM) affiliated to the University of Montreal. He is also the Research Integrity and Ethics Officer for the IRCM. Pr. Coulombe has published 106 scientific articles so far (totalling nearly 9000 citations) and was among the pioneers in applying AI to protein interactomes (see Jeronimo et al 2007). He is the founder of LivingMind SuperAI, now LivingMind AI, and a strong advocate for life-aligned technological futures. 

LINKEDIN ACCOUNT: 

www.linkedin.com/in/benoit-coulombe-31a22a132

The LivingMind AI framework is governed by five general principles, enforced by six specific pillars representing practical applications. 

PRINCIPLES

1. Biological Truth : Intelligence grounded in living processes from cells to ecosystems.
2. Ethical Integrity : Safety/ethics integrated into      architecture and training objectives.
3. Global Equity : Access, benefit-sharing, and anti-concentration measures.
4. Ecological Balance : Compliance with planetary boundaries.
5. Existential Care : Orientation to long-term human and biospheric flourishing.

PILLARS

• LivingMap: A dynamic cartography of intelligences (biological, artificial, hybrid), their uses, risks, and interdependencies.
• LifeSentinel: A risk-radar + guardrail layer (red-teaming, incident reporting, safety dashboards, user-level controls).
• LifeIndex: Metrics linking AI actions to human well-being, ecological impact, and equity outcomes.
• LivingAgora: Participatory governance (multi-stakeholder deliberation, audits, standards, escalation pathways).
• LifeLearning: Continuous, open scientific + civic learning loop (benchmarks, incident      corpora, reproducibility).
• Interactys-AI: The Interactys-AI initiative is aimed at identifying protein-protein interactions for drug and antiviral discovery.

Overview: Protein-protein interactions, collectively called Interactome, are essential to sustain the function of living cells and faulty interactions can cause disease. They are also involved in virus infections, making some of them ideal targets for antiviral discovery.

PROJECT #1

Joint Special Issue of the Journals Biomolecules (IF 4.8) and Viruses (IF 3.5)

Topic: Virus-Host Protein Interactions

Guest Editor

Prof. Benoit Coulombe
1. Translational Proteomics Laboratory, Institut de Recherches Cliniques de Montréal, Montréal, QC H2W 1R7, Canada
2. Department of Biochemistry and Molecular Medicine, Université de Montréal, Montréal, QC H3T 1J4, Canada

Dear Colleagues,

Protein–protein interactions play essential roles in viral infections. Some viral proteins join together to form protein complexes that exert key roles in the virus life cycle. The RNA polymerase complex is an example. Additionally, some viral proteins interact with host factors during the infection process, usurping the function of these factors in favor of viral functions. Together, these protein–protein interactions are central in the pathophysiology of viral infection. As a consequence, perturbation of these key interactions has the potential to impair virus replication and propagation. Over recent years, a number of molecules, including small chemicals, peptides and antibodies, have been reported to interfere with viral functions by perturbating protein–protein interactions. The molecules showing virus specificity can serve as antiviral agents en route to the discovery of new drugs. This joint Special Issue of the open access journals Viruses (Impact Factor 3,5) and Biomolecules (Impact Factor 4,8) is dedicated to experimental studies or reviews dealing with the use of protein–protein interactions as targets for antiviral drug discovery. This Special Issue also represents a powerful tool in the combat against threatening human viruses.

Manuscript Submission Information on the MDPI Website - https://www.mdpi.com 

• Ease of navigation: Grouping papers by topic helps scholars navigate broad scope journals more efficiently.
• Greater discoverability: Special Issues support the reach and impact of scientific research. Articles in Special Issues are more discoverable and cited more frequently.
• Expansion of research network: Special Issues facilitate connections among authors, fostering scientific collaborations.
• External promotion: Articles in Special Issues are often promoted through the journal's social media, increasing their visibility.
• Reprint: MDPI Books provides the opportunity to republish successful Special Issues in book format, both online and in print.

PROJECT #2

H5N1 protein interactions with human factors (data available soon)

H5N1 is a highly pathogenic avian influenza. It is tightly monitored since 2023-2025 because there has been intense circulation among wild and farmed birds, with occasional spillover into humans as cases were reported in Chile, the UK, and the USA. The pandemic risk is HIGH because it is already transmitted to mammals (minks, seals, cattle) and continues to acquire mutations that could facilitate human-to-human transmission. The emergence of a variant that can spread efficiently between humans could be catastrophic. We currently develop and use AI procedures to identify host factors that are targeted by H5N1 proteins as they can lead to antiviral discovery.

Moursli Y, Faubert D, Grou C, Coulombe B. Discovery and characterization of a pancreatic β cell subpopulation expressing an unknown surface epitope through single cell proteomics. 2025 bioRxiv. https://lnkd.in/e8GDAs5i 

Moursli Y, Poitras C, Coulombe B. Investigating pancreatic β cell membrane epitopes using unbiased cell-based Fab-phage display. bioRxiv, 2025.03.29.645157; doi: https://doi.org/10.1101/2025.03.29.645157. 

Hashimoto-Roth E, Forget D, Gaspar VP, Bennett SAL, Gauthier MS, Coulombe B, Lavallée-Adam M. MAGPIE: A Machine Learning Approach to Decipher Protein-Protein Interactions in Human Plasma. J Proteome Res. 2025 Jan 7. doi: 10.1021/acs.jproteome.4c00160. Epub ahead of print. PMID: 39772751.

Pinard M, Dastpeyman S, Poitras C, Bernard G, Gauthier MS, Coulombe B. Riluzole partially restores RNA polymerase III complex assembly in cells expressing the leukodystrophy-causative variant POLR3B R103H. Mol Brain. 2022 Nov 30;15(1):98. doi: 10.1186/s13041-022-00974-z

Sokullu E, Pinard M, Gauthier MS, Coulombe B. Analysis of the SARS-CoV-2-host protein interaction network reveals new biology and drug candidates: focus on the spike surface glycoprotein and RNA polymerase. Expert Opin Drug Discov. 2021 

Aug;16(8):881-895. doi: 10.1080/17460441.2021.1909566.

Sokullu E, Gauthier MS, Coulombe B. Discovery of Antivirals Using Phage Display. Viruses. 2021 Jun 10;13(6):1120. doi: 10.3390/v13061120. 

Cloutier P, Poitras C, Durand M, Hekmat O, Fiola-Masson É, Bouchard A, Faubert D, Chabot B, Coulombe B. R2TP/Prefoldin-like component RUVBL1/RUVBL2 directly interacts with ZNHIT2 to regulate assembly of U5 small nuclear ribonucleoprotein Nat Commun. 2017 May 31;8:15615. doi: 10.1038/ncomms15615. PMID: 28561026

Houry WA, Bertrand E, Coulombe B. The PAQosome, an R2TP-Based Chaperone for Quaternary Structure Formation Trends Biochem Sci. 2018 Jan;43(1):4-9. doi: 10.1016/j.tibs.2017.11.001. Epub 2017 Dec 5. Review. PMID: 29203338

Thiffault I, Wolf NI, Forget D, Guerrero K, Tran LT, Choquet K, Lavallée-Adam M, Poitras C, Brais B, Yoon G, Sztriha L, Webster RI, Timmann D, van de Warrenburg BP, Seeger J, Zimmermann A, Máté A, Goizet C, Fung E, van der Knaap MS, Fribourg S, Vanderver A, Simons C, Taft RJ, Yates JR 3rd, Coulombe B, Bernard G. Recessive mutations in POLR1C cause a leukodystrophy by impairing biogenesis of RNA polymerase III Nat Commun. 2015 Jul 7;6:7623. doi: 10.1038/ncomms8623. PMID: 26151409

Jeronimo C, Forget D, Bouchard A, Li Q, Chua G, Poitras C, Thérien C, Bergeron D, Bourassa S, Greenblatt J, Chabot B, Poirier GG, Hughes TR, Blanchette M, Price DH, Coulombe B. Systematic analysis of the protein interaction network for the human transcription machinery reveals the identity of the 7SK capping enzyme Mol Cell. 2007 Jul 20;27(2):262-74. PMID: 17643375

LivingMind AI: Toward a Life-Aligned Artificial Intelligence Architecture

Advanced AI is entering a phase where raw capability expansion must be complemented by biological grounding, ethical integration, and measurable societal trust. Recent policy and technical developments, such as OpenAI’s product-embedded teen-safety mechanisms and the emergence of enforceable safe-by-design standards via LawZero, highlight a shift from post-hoc control to trust-by-design.

At the same time, frontier bioscience AI (e.g., AlphaFold 3 and successor multi-omics models) demonstrates both unprecedented therapeutic potential and profound dual-use concerns, underscoring the urgency of frameworks that integrate capability, governance, and biological constraints.

We introduce LivingMind AI, a life-aligned intelligence architecture grounded in five principles (Biological Truth, Ethical Integrity, Global Equity, Ecological Balance, Existential Care) and operationalized via six implementation layers: LivingMap, LifeSentinel, LifeIndex, LivingAgora, LifeLearning, and the Interactys-AI biomedical program.

Unlike risk-centric models focused primarily on adversarial control, LivingMind AI treats life alignment as a systems property, combining biological heuristics, multi-stakeholder governance, transparent auditability, and ecological accounting. Central to this approach is the LifeIndex, a suite of measurable indicators quantifying AI’s contribution or harm across human health, planetary stability, and equitable access.

We argue that AI must evolve not above life, but within life’s boundaries, dynamics, and ethics. LivingMind AI offers a scaffolding for future standards, research agendas, and translational programs aligned with long-term biospheric and societal flourishing.