Back
Maybe Anthropic's Long-Term Benefit Trust is powerless
webAuthor
Zach Stein-Perlman
Credibility Rating
3/5
Good(3)Good quality. Reputable source with community review or editorial standards, but less rigorous than peer-reviewed venues.
Rating inherited from publication venue: LessWrong
Relevant to researchers and policymakers evaluating AI lab governance structures; raises questions about whether safety-oriented corporate mechanisms can withstand investor pressure at major frontier AI labs.
Forum Post Details
Karma
206
Comments
21
Forum
lesswrong
Forum Tags
Anthropic (org)AI GovernanceAI
Metadata
Importance: 62/100blog postanalysis
Summary
Zach Stein-Perlman critically examines Anthropic's Long-Term Benefit Trust, arguing it may be effectively powerless because stockholders can override or dissolve it via supermajority votes with undisclosed thresholds. The analysis highlights that major investors like Amazon and Google could potentially constitute such a supermajority, and Anthropic's refusal to publish the Trust Agreement suggests the details would reveal its weakness.
Key Points
- •Anthropic's Long-Term Benefit Trust can apparently be overridden, modified, or abrogated by a supermajority of stockholders without Trustee consent.
- •The supermajority thresholds are undisclosed, making it impossible to assess how easily major investors like Amazon or Google could override the Trust.
- •Even if stockholders never formally override the Trust, the implicit threat may prevent Trustees from acting against stockholder interests.
- •Anthropic has not published the Trust Agreement, and the author argues this opacity likely benefits Anthropic by obscuring the Trust's weaknesses.
- •Key transparency gaps include undisclosed board membership, unconfirmed Trust elections, and unclear enforcement mechanisms.
Cited by 1 page
| Page | Type | Quality |
|---|---|---|
| Anthropic Long-Term Benefit Trust | Organization | 70.0 |
Cached Content Preview
HTTP 200Fetched Mar 15, 202618 KB
x This website requires javascript to properly function. Consider activating javascript to get access to all site functionality. Maybe Anthropic's Long-Term Benefit Trust is powerless — LessWrong Anthropic (org) AI Governance AI Personal Blog 206
Maybe Anthropic's Long-Term Benefit Trust is powerless
by Zach Stein-Perlman 27th May 2024 3 min read 21 206
Crossposted from AI Lab Watch. Subscribe on Substack.
Introduction
Anthropic has an unconventional governance mechanism: an independent "Long-Term Benefit Trust" elects some of its board. Anthropic sometimes emphasizes that the Trust is an experiment, but mostly points to it to argue that Anthropic will be able to promote safety and benefit-sharing over profit. [1]
But the Trust's details have not been published and some information Anthropic has shared is concerning. In particular, Anthropic's stockholders can apparently overrule, modify, or abrogate the Trust, and the details are unclear.
Anthropic has not publicly demonstrated that the Trust would be able to actually do anything that stockholders don't like.
The facts
There are three sources of public information on the Trust:
The Long-Term Benefit Trust (Anthropic 2023)
Anthropic Long-Term Benefit Trust (Morley et al. 2023)
The $1 billion gamble to ensure AI doesn't destroy humanity (Vox: Matthews 2023)
They say there's a new class of stock, held by the Trust/Trustees. This stock allows the Trust to elect some board members and will allow them to elect a majority of the board by 2027.
But:
Morley et al.: "the Trust Agreement also authorizes the Trust to be enforced by the company and by groups of the company’s stockholders who have held a sufficient percentage of the company’s equity for a sufficient period of time," rather than the Trustees. I don't know what this means.
Morley et al.: the Trust and its powers can be amended "by a supermajority of stockholders. . . . [This] operates as a kind of failsafe against the actions of the Voting Trustees and safeguards the interests of stockholders." Anthropic: "the Trust and its powers [can be changed] without the consent of the Trustees if sufficiently large supermajorities of the stockholders agree." It's impossible to assess this "failsafe" without knowing the thresholds for these "supermajorities." Also, a small number of investors—currently, perhaps Amazon and Google—may control a large fraction of shares. It may be easy for profit-motivated investors to reach a supermajority.
Maybe there are other issues with the Trust Agreement — we can't see it and so can't know.
Vox: the Trust "will elect a fifth member of the board this fall," viz. Fall 2023. Anthropic has not said whether that happened nor who is on the board these days (nor who is on the Trust these days).
Conclusion
Public information is consistent with the Trust being quite subordinate to stockholders, likely to lose their powers if they do
... (truncated, 18 KB total)Resource ID:
2108eeb7e0a8bd62 | Stable ID: NjIyYzIxZm