Driving Responsible Innovation of AI, Life Sciences and Next Generation Biotech

Array

We are making God obsolete,” declared the rockstar-like professor of synthetic biology in the recent Netflix hit show Biohackers. What happens when humans become creators of new forms of DNA, including our own? Who gets to decide? Who owns our genomic data? Who is it shared with? Who is accountable? To what end? Just how far are humans willing to go in the name of science, progress, and competitive advantage in the biotechnology twilight zone?

Modifications of the human genome have long been a concern in scientific communities.

Five years ago, a group of 18 scientists and ethicists, including Jennifer Doudna, winner of the 2020 Nobel Prize in chemistry, warned that CRISPR, a revolutionary genomic scissor used to cut and splice DNA, should be used cautiously when attempting to correct human genetic diseases. The scientists strongly discouraged any attempts at making deliberate changes to human germline cells whose genetic material gets passed on to the next generation.

The same year, more than 1,000 experts, scientists, and researchers warned about the dangers of artificial intelligence (AI) becoming weaponized, resulting in autonomous kills chains and weapon systems. The uncomfortable reality of both these noble initiatives, however, is that potential harms from these technologies will likely not be purpose-built, but rather derived from repurposing existing, commercial, and industry-driven technologies.

The Human Genome Project, the international collaborative effort to map every gene in the human genome, launched in 1990, remains one of the biggest scientific collaborations ever undertaken. Today, thanks to AI, mapping a human genome or tracing your ancestral heritage may take less than a day and costs are rapidly dropping.

“Expedited by COVID-19, there has been a proliferation of actors, technologies, and methodologies featuring AI and the life sciences.”

Comparing DNA sequences can shine a light on predisposition to diseases, may diagnose certain cancers, identify antibodies, or guide patient treatment. Such technologies can also be used, as we have seen recently, to identify strains of viruses, track their evolution, and even help to identify their source. COVID-19 has provided the impetus for initiatives to share viral genomes globally, to facilitate genetic collection and accelerate genomic profiling. Expedited by COVID-19, there has been a proliferation of actors, technologies, and methodologies featuring AI and the life sciences.

Beyond Privacy Considerations

The transformative impact of these technologies and the commodification of our biological and genomic data continue to change how we view health and treat disease, and will have a significant impact on the future biological continuum and geopolitical order. Traditional methods of minimizing harm from misuse, dual use, and hostile use are increasingly at risk of colliding with the complexity of algorithmic functions and geostrategic competition.

Developments in bioinformatics coupled with machine learning have potential to be utilized for coercive or lethal purposes. New and intrusive forms of surveillance fuelled by personal and biological data and the commodification of our bio data flows, prompted by a global health crisis such as the COVID-19 pandemic, can enable unprecedented social, biological, and consumer control.

Genetic data has become a valuable data trove.

For instance, in the midst of the pandemic, Blackstone, a private equity company, recently acquired the genealogy and genetic heritage provider Ancestry, known for its “pop genetics” to map genetic heritage, with more than 18 million users, for 4.7 billion dollars. The big technology giants have similarly made moves to gain access to personal health information to both develop products to transform the health industry as well as gain access to bio data to transform the next iteration of AI developments.

Alphabet, the multinational conglomerate and parent company of Google Inc. pivoted into predictive and personalized health offerings. Google recently partnered with Ascension, a health systems provider, to collect data to design hyper-personalized AI-powered health care services.

In the United Kingdom, another Alphabet subsidiary, DeepMind, made headlines and attracted public scrutiny using data from the UK’s National Health Service. Palantir Technologies, which is building principal software platforms that integrate sensitive data to perform specialized sequencing, is another actor in the new health data market.

“The transformative impact of these technologies and the commodification of our biological and genomic data continue to change how we view health and treat disease, and will have a significant impact on the future biological continuum and geopolitical order.”

As governments and companies are under pressure from COVID-19 to transform their information technology infrastructure, due diligence is seemingly glossed over. In late 2019, the US Department of Defense warned its military personnel not to participate in direct-to-consumer genetic testing, citing “personal and operational” security concerns. And the UK established an independent biosecurity centre to monitor virus related threats.

New app-based contact-tracing tools have been created, some more haphazardly than others, often at a great financial cost. This has detracted from investments in more traditional and proven manual contact-tracing systems and human system upgrades, leading the Council of Europe to question if the promises made about these apps are “worth the predictable societal and legal risks.”

Beyond privacy considerations, it is clear that our immutable, biological selves are no longer sacrosanct. Exploiting biological data using computational power also presents challenges for biosecurity frameworks, in particular, those geared towards preventing accidental or deliberate releases of pathogens.

In an essay on averting the hostile exploitation of biotechnology, Matthew Meselson, professor of molecular biology at Harvard University and one of the pioneers of biological arms control, posited in 2000 that “as our ability to modify fundamental life processes continues its rapid advance, we will be able not only to devise additional ways to destroy life but will also become able to manipulate it—including the processes of cognition, development, reproduction and inheritance […] Therein could lie unprecedented opportunities for violence, coercion, repression, or subjugation.”

The current biological arms control regime, the Biological Weapons Convention, has been in force since 1975. The treaty comprehensively prohibits biological weapons, understood as biological agents used for harmful purposes, and member countries agree that it unequivocally covers all microbial or other biological agents or toxins, naturally or artificially created or altered, as well as their components, whatever their origin or method of production.

Yet, advances in genomics and the life sciences, coupled with the capabilities from unprecedented computational power and sophisticated health and bio data-gathering tools, have clearly demonstrated the limitations of this and related regimes to address harm caused not by biological agents, but by intercepting or interfering with biological processes.

Managing Risks of AI in Life Sciences

The window of opportunity to minimize risks to humanity, security, and stability posed by advances in AI and biology is quickly closing. Advances in computational biology and collections of our genetic data are moving at warp speed. We need to urgently build networks, knowledge, and capacity to contemplate and respond to these risks.

“The window of opportunity to minimize risks to humanity, security, and stability posed by advances in AI and biology is quickly closing. Advances in computational biology and collections of our genetic data are moving at warp speed. We need to urgently build networks, knowledge, and capacity to contemplate and respond to these risks.”

As a first step, efforts should be directed at the intersection between the biological disarmament community, numerous AI fora, and the global standards community. These efforts should focus particularly on building norms and control systems that enable transparency, provide insight on the intent behind dual-use activities, and foster trust. In addition, they must support traditional arms control initiatives to build confidence like peer reviews, scientific collaboration, and expert-level exchanges. The challenges of converging technologies will require bold ideas and new relationships to re-envision responsible bio innovation for the future.

Managing the fast and broad technological advances now underway will require new governance structures that draw on individuals and groups with cross-sectoral expertise— from business and academia to politics and defense. These networks or communities of practice will contribute to identify emerging security risks and make recommendations for dealing with them and to enable and empower these communities to speak directly with each other in order to manage and minimize potential harms and risks stemming from AI and the life sciences.

The lack of a collaborative platform and a regular technical and scientific dialogue impact the ability to comprehensively and collectively review and appropriately address the risks arising from genomic technologies, bioinformatics, and machine learning. A lack of shared scientific and technical standards, as well as the lack of a common vernacular to define the qualities of these transformative technologies and potentially bio-invasive surveillance tools, raise multiple concerns and require urgent attention. The proliferation of data-reliant health technologies, the fast-growing biotechnology community, and the ubiquity of health data will likely contribute to the onset of new biometric-based revenue streams and the onset of a new bio-digital age.

The lack of global and shared industry standards in key domains of the life sciences with a deep and profound impact on international security not only hinders the interoperability of such technologies, but the void also raises a separate set of concerns, which complicate the assessment of opportunities and regulatory compliance. There have not yet been specific consortia or associations in this area directly focused on the development of official standards or working with the standards ecosystem. Although some of the ISO standards under negotiation will apply, none specifically contemplate the merging of AI and the life sciences.

It is time to respond to the risks of increasingly fractured bioinformatics policies and standards amidst a growing bio data economy and unprecedented amounts of bio data flows and bio-algorithms.

“It is time to respond to the risks of increasingly fractured bioinformatics policies and standards amidst a growing bio data economy and unprecedented amounts of bio data flows and bio-algorithms.”

New IEEE SA Industry Connections Program

The IEEE Standards Association (IEEE SA) launched an Industry Connections (IC) Program IC20-027- Responsible Innovation of AI and the Life Sciences, bringing together exceptional individuals from business, academia, politics, defense, civil society, and international organizations, to oversee developments relevant to biological threats in science, business, defense, and politics and to advise on concerted cross-sector actions.

No individual, discipline, forum, or institution can adequately respond to the threats the world now faces – it will take a chorus of voices listening to each other, and a creative community to continue to address these risks in a meaningful way. Doing so will have profound impacts on the future of our planet.

Resources

Authors:

  • Nancy D. Connell PhD, IC20-027 Industry Connections (IC) Program Co-Chair
  • Kobi Leins, IC20-027 Industry Connections Program (IC) Co-Chair
  • Filippa Lentzos PhD, IC20-027 Industry Connections (IC) Program Co-Chair
  • Anja Kaspersen, IC20-027 Industry Connections Program (IC) Co-Chair

Share this Article