VERITAS: Verifiable Accountability For Autonomous AI Agents

Florida, USA

Cryptographically signed, independently verifiable receipts for what AI agents actually did, anchored to Bitcoin so the record can't be quietly rewritten.

Donate:Manifund

Cryptographically signed, independently verifiable receipts for what AI agents actually did, anchored to Bitcoin so the record can't be quietly rewritten.

Donate:Manifund

Project Details

My project, VERITAS, makes AI prove what it did on the Bitcoin mainnet. It does this by producing tamper-evident, independently verifiable receipts for AI inferences and agent actions that become anchored to Bitcoin. In doing so, the record can't be quietly rewritten after the fact. Veritas is already built publicly with 10 MIT-license repositories and has multiple real Bitcoin mainnet anchors.

I want to take VERITAS's agent-accountability layer from working prototype to a tool that real AI operators and outside auditors use to check what an autonomous agent actually did, without having to trust the operator, making this upcoming generation of implemented AI tools trustworthy to the public.

Here's Where Things Stand Now: VERITAS can have your agent sign its action(s) on the Bitcoin Mainnet locally through the use of cryptography. Every local record that's attested can be proven genuine since VERITAS contains a live verifier repository.

What Results Will Yield Exactly From Payment? I need capital to handle two things that'll be very beneficial to both VERITAS, as a project, and the autonomous AI space, which will be the necessary essentials that's going to help with publishing on the public relay pipeline via Bitcoin anchoring for the creation of my third-party auditing tool to take an agent's very own public key as input, which retrieves the records from the agent once they're published under it. This is a concrete follow-the-money method that'll allow agents to pull entire histories with verified receipts of actions or the lack thereof.

How Can We Tell What Worked Or Didn't Work? The whole point of VERITAS is maintaining verification of an agent's record history so if I were to check for 100 anchored records on the mainnet chain but the operator only manages to reveal 90 records from the public relay to me with flagged dates, the mismatch of records is immediately noticeable in this situation.

I would hope to not only get the necessary funds for the minimal build but to get VERITAS to its fullest capacity in later stages of the project, which also include installing a live operator to help maintain records.

Theory of Impact

I believe verifiable AI is going to be a matter of justice in the near future, AI activity is already being entered as evidence in serious matters and how things get decided. That's a real fear for the public, nobody actually has a say-so in what we're forced to accept as proof from someone else's word. In a world of autonomous AI agents, VERITAS adaptation reduces high levels of suspicion by using a neutral source that any person and any machine are able to check themselves while also holding the actions of an agent accountable.

People

Tutankhamun Castillo El-Bey

Founder & Director

Funding Details

Start Date: May 23, 2026
End Date: -
Expected Duration: Ongoing; current funded roadmap covers 12 months
Funding Raised to Date: -
Annual Budget: -
Monthly Burn Rate: -
Current Runway: -
Funding Goal: -
Funding Stage: Seeking first grant; self-funded to date
Fiscal Sponsor: -

Track Record

2026, all public and independently checkable:

May: VERITAS open-source ecosystem published. 10 MIT-licensed repositories, 660+ passing tests, formal RFC-style specification with canonical test vectors (github.com/Ifasola34).

June 13: first Bitcoin mainnet anchor (tx 92b2c4e434ae...215aafa0, verifiable on mempool.space).

June 16: second mainnet anchor (tx 5f46b8c09980...7a7afadb); salted commit-and-reveal privacy layer shipped.

June: public in-browser verifier hardened (vrt1-web-verifier.pages.dev); a full verification runs in about two minutes with no install and no account.

June 24: book-length design document published: Proof of Agent (Kindle and paperback).

Grants Received– no grants recorded

Add grant

Funding Asks

grantmaking.ai Launch Round

Applied

Minimum

$25,000

Ideal

$100,000

What this ask is

How the money will be spent

- $25,000 (months 1 to 3), the minimum: harden vrt1-agents to production and ship the third-party auditing tool.

$50,000 (months 4 to 6): add the first live integration with a real agent operator, demonstrating end-to-end third-party verification.
$75,000 (months 7 to 9): add the two safety benchmarks (provenance-laundering and agent-to-agent forgery detection) and an independent external security audit.
$100,000 (months 10 to 12), the full goal: a second integration with a different operator, a second audit round, and a public technical report.

Composition of the full $100,000 (the same line items as my other applications, scaled to twelve months):
$60,000, my time (project lead, 12 months; gross, including self-employment and income tax), covering the four quarterly work blocks above.
$15,000, a contract cryptography/security engineer to co-build and harden the verification path.
$13,000, an independent external security audit by a separate reviewer, with findings published.
$4,000, compute and infrastructure: Bitcoin anchoring fees, CI, hosting, and benchmark compute.
$8,000, buffer for unplanned costs.

Discussion

No comments yet. Be the first to share your thoughts.