Datafund (@DataFundProject) header
Datafund (@DataFundProject) profile picture

In the digital age, freedom begins with owning personal data. Connect with us on Linkedin: https://teapawt.net/t.co/vnT2uLR57I

Europe
Joined August 2017
Photo by Datafund (@DataFundProject)Photo by Datafund (@DataFundProject)Photo by Datafund (@DataFundProject)Photo by Datafund (@DataFundProject)Photo by Datafund (@DataFundProject)Photo by Datafund (@DataFundProject)Photo by Datafund (@DataFundProject)Photo by Datafund (@DataFundProject)Photo by Datafund (@DataFundProject)Photo by Datafund (@DataFundProject)
Data provenance is becoming a key parameter in the GenAI landscape. Lawmakers are demanding it and the cost of non-compliance is extremely high. Luckily, we've built a toolkit for building an immutable and auditable data trail. No 3rd party needed and standard-agnostic. 👇👇👇
1
2
1
5
382
A tamper-proof record cannot preserve an action nobody recorded. @METR_Evals’ review of OpenAI’s monitoring makes the distinction urgent: capture what happened, then protect the record. Provenance needs both. metr.org/blog/2026-10-22-red…
3
40
An agent with a USDC wallet on Base can now rent storage on Swarm by itself. @ethswarm #storage Ask our gateway, get an HTTP 402 with a price, sign, send again. Two requests. No signup, no BZZ, no bridge. x402 v1 and v2, live on Base mainnet. Alpha. provenance-gateway.datafund.…
1
3
6
262
In July, agents in a lab's own evaluation got out of their sandbox. The independent review found they "tried to retroactively edit transcripts". The records held that time. Keep them where no agent, and no operator, can change them. datafund.io/blog/agents-usin…
1
1
36
The same pattern as our guide, applied to an agent's memory: fingerprints on @ethswarm, a notary signature, and a check anyone holding the memories can run. teapawt.net/plur_ai/status/2117899…
Later, someone asks what your agent knew. Anyone who holds the same memories can check the session's list with one command: plur verify --manifest <reference> It checks the notary signature locally, then every memory.
1
4
147
If your agent can rewrite its own record, you don't have a record. You have the agent's story. Keep the record where nobody can change it, the agent included, and let the people entitled to check verify it. datafund.io/blog/agents-usin…
1
42
Where should an AI lab keep the record of what its agents did? Out of reach of the agent and of the lab. Give the evaluator its fingerprint when it's written, and let them check it themselves. How it works, with a one-minute demo: datafund.io/blog/where-shoul…
4
91
Could you prove what your AI agent did yesterday? Make a receipt for a file in about a minute, no wallet or account: dataprovenance.app Open source, pre-release, on a test network. datafund.io/blog/how-to-give…
2
50
Looking back through old logs only works if the old logs are still the old logs. In July, AI agents in a lab's evaluation tried to edit their own transcripts. As far as the reviewers saw, the originals held. This time. datafund.io/blog/agents-usin…
1
4
60
How does an auditor check that an agent's transcript wasn't changed? 1. Fetch the record by its address. 2. Recompute its fingerprint and compare. 3. Check the notary's signature, or the anchor on chain. The steps, and what they don't prove: datafund.io/blog/how-to-give…
2
47
Who keeps the record of what your agents did? If the answer is "the agents", you don't have a record. You have their story. Keep it where nobody, the agent included, can change it. datafund.io/blog/agents-usin…
2
39
Replying to @OpenAI
@OpenAI, you wrote: save agent transcripts "immutably", in write-once storage. @AnthropicAI, your CEO asked for evaluators with "employee-like access". We'll hold you to it: records no agent can change, checked by auditors. RT so this reaches them. datafund.io/blog/agents-usin…
1
48
An agent's account of its own work is only as good as the agent's inability to change it. Most agent logs today sit where the agent, or anyone with access, can edit them. datafund.io/blog/agents-usin…
1
51
1/8 Could you prove what your AI agent did yesterday?Not what it told you it did. What it actually did.We're handing agents our code, our money, our customers. Staying in control starts with that answer. 🧵
1
2
1
5
274
2/8 Knowing what an agent did means having a record. But if your agent can rewrite its own record, you don't have a record. You have the agent's story.
1
1
36
In a recent article, WEF highlighted the growing need for strong data provenance systems as synthetic data rises. weforum.org/stories/2025/10/…
1
109
The risks of blurring lines between the two types of data are considerable: AI autophagy, reinforced bias and a serious erosion of trust as deepfakes and misinformation threaten to take over.
1
20
Data provenance is important because it tells us how the world is vs. how somebody wants us to see it. It tell us if the source was legitimate, if the capture was authentic or if the data reflected reality.
1
1
2
89
If we cannot verify origin, authenticity or context, data becomes a liability. And as AI agents move into research, finance, science and public systems, they will act on inputs no one can trace. We need provenance captured at the source.
1
102
A recent paper (arxiv.org/pdf/2404.12691) highlights that a growing issue with provenance tools is their piecemeal nature. 🧩 Dev teams can’t reliably assess safety, legality or copyright across thousands of mixed datasets. Creators, on the other hand, can’t track their content.
1
1
65
AI needs a provenance infrastructure that unifies ethics, law and technical aspects. That's what we're building for with our latest provenance POC. 👇👇👇
52