In the realm of AI discourse, much attention is directed towards fostering responsible development of the technology itself (e.g. Is it free from bias? Transparent? Explainable?).  Whether or not AI is developed responsibly, it functions as an accelerant; it can be either a clean or dirty one. 

Beyond these technical considerations lies a crucial question: what are we accelerating, and what is the underlying business model?  How does the company creating the technology think about the spread of value that it creates, destroys, or extracts from various stakeholders? Who is it incentivized to create value for?  How will it be governed?  

This month, three major foundations —  Ford Foundation, Omidyar Network, and Nathan Cummings Foundation— jointly announced their purchase of shares in Anthropic PBC, a competitor to OpenAI, widely considered the industry leader.

Anthropic’s emergence comes amidst widespread criticism aimed at OpenAI for allegedly veering away from its original mission of benefiting humanity. In the wake of drama between OpenAI’s now-former board members and once-again CEO Sam Altman, the critique has spanned various actors and sectors— from civil society to Elon Musk.

While venture-backed companies primarily serve the interests of their shareholders, the far-reaching implications of AI necessitate broader consideration of stakeholders. Impact investors, foundation endowments, and university endowments have a role to play in democratizing AI, and keeping it in check, by advocating for a more diverse representation of interests within a company’s capital structure.

Responsible governance 

Civil society has honed its ability to influence public companies through external advocacy, activism, and policy. Now we get to see how it might wield influence as investors, especially when investing alongside major shareholders like Amazon, which has invested $4 billion in Anthropic PBC.

While Anthropic PBC may not be an “impact company” in the traditional sense, like many other cutting edge technology companies, it is poised to disproportionately influence the playing field of impact. It  has also gestured towards a next-generation ethos that rejects the “growth at all costs” mentality that has dominated Silicon Valley and has been at the heart of OpenAI controversy.  

Anthropic has gone so far as to incorporate as a Public Benefit Corporation, which legally codifies its commitment to a public benefit purpose beyond generating profit, by “allowing its directors to balance the financial interests of the stockholders with the public benefit purpose specified in the corporation’s certificate of incorporation.”  

Anthropic alsohas  established a Long Term Benefit Trust to help ensure integrity of its mission as it grows. The trust “is an independent body of five financially disinterested members with an authority to select and remove a portion of our Board that will grow over time.”  

While the collective investment of the Ford, Omidyar, and Nathan Cummings Foundations is limited in size and does not afford specific governance or voting rights, we can’t help but wonder if there is an opportunity for these foundations to punch above their weight in terms of influence. From their position as shareholders, can they more effectively educate or inspire Anthropic to consider new paradigms of value when it comes to human rights, the future of work, or access to essential services?  

As an AI company with Amazon as a major investor, for example, technology being used in ways that threaten labor rights ought to be a real concern. Strategic guidance from experienced entities like Ford could help mitigate negative downstream effects, or encourage and enable the development of Anthropic PBC’s technology in other ways that might better serve vulnerable constituencies.

To support responsible innovation, we believe there is a need for more impact capital that can make early stage investments in transformative technology companies that affect the playing field of impact, even if the companies are not expressly “impact” themselves.

“Given that philanthropic capital is less than a quarter of a percent of the capital markets, I believe that foundations, philanthropies and impact investors should be looking at opportunities in the for-profit world, like PBCs with real mission lock, to accelerate positive impact and mitigate negative outcomes associated with the development and wide-spread adoption of new technologies,” says Susan Mac Cormac, head of sustainability practice at Morrison Foerster.

Mac Cormac, a corporate governance expert, stresses the importance of philanthropic capital engaging with for-profit entities, to drive positive impact and mitigate negative outcomes associated with technological advancements. The key is governance provisions that ensure social capital has a voice at the board level, along with “mission-lock” structures to secure the public benefit even through dilution, growth and exits.

”The presence of these impact investors on Anthropic PBC’s cap table is an important signal, leaning into the power of investors to drive impact through voice and governance,” said NYU’s Andrea Armeni, co-author of “Alternative Ownership Enterprises.”

Armeni pointed to Omidyar Network’s focus on “efforts to promote and support new corporate governance formats that reflect the interests of multiple stakeholders, and not just financial shareholders.” 

“This is a prime example of getting high impact leverage from a relatively small stake: these investors can wield an outsized voice and ensure that the commitment in principle baked into a Public Benefit Corporation structure is actually carried out,” Armeni said. “This is an explicit backing of an alternative model that takes governance seriously.”

Leaving impact on the table

The culture and incentive structures of technology companies become less pliable as they mature, so the earlier the investment, the greater leverage to influence how a company grows.  

We argued in SSIR that “Impact investors must become exponentially more active at the seed and startup stages,” and that “matching Y Combinator’s speed and size, and learning from their playbook on de-risking early-stage investing at scale to help influence the ethos with which tech companies are encouraged to grow.“

Ideally, investors would “both commit capital to funders who are exploring broader governance and ownership models, and wield their power to spur others to go down that path, whether through direct influence or by setting examples,” adds Armeni.  We know from our experiences working with founders at Y Combinator, Startups & Society and Responsible Innovation Labs, as well as teaching aspiring entrepreneurs at Cornell Tech and MIT, that many aspiring tech founders recognize the flaws of a growth-at-all-costs mentality.  

They aspire to build companies that prioritize holistic value creation. They want to account for unintended consequences and balance the interests of all of their stakeholders. Yet often they don’t identify explicitly as ‘Impact Founders’ and struggle to find capital that truly values responsible innovation.

These founders require robust capital infrastructure to support their endeavors. While Anthropic PBC represents a promising start, it is one company in a single vertical. Hopefully, it marks the beginning of a broader mobilization of impact capital towards more responsible and impactful technology development.

Lyel Resner is a visiting faculty and head of the public interest technology studio at Cornell Tech.

Dr. Wilneida Negrón is a strategic advisor at Responsible Innovation Labs.