MACHINE LEARNING LAB S.R.O.

• SanctionsRecords foundSanctioned by 1 authority: OFAC. Confirmed present in a source snapshot within the last 90 days.
• Debarment & exclusion— Checked, nothing found16 sources read, none holding a record for this company.
• Ownership & controlRecords found1 designated individual recorded in connection with this company: KLIMINA, Yevgeniya.
• People— Not assessedNothing has been read for this company on this dimension.
• Identity & registration— Not assessedNothing has been read for this company on this dimension.
• Jurisdiction— Not scoredNot scored. The jurisdiction lists this system reads are: EU list of high-risk third countries (AMLD Art. 9). Each names specific countries, so absence from them is not a clearance of Czechia. Separately, 6 of the sanctions authorities this system reads have designated companies registered in Czechia — a count of this corpus, not a rating of the jurisdiction.
• Court records— Not assessedNothing has been read for this company on this dimension.

• OFACActive

  RUSSIA-EO14024

  Designation date not published by this authority · OFAC SDN

No corporate ownership records on file for this company. Ownership is recorded only where a register or a designating authority publishes it, and coverage is thin for companies outside the jurisdictions with open beneficial-ownership registers.

Legal name

MACHINE LEARNING LAB S.R.O.

Also known as

MACHINE LEARNING LAB S.R.O.

Jurisdiction

CZ

First recorded here

18 Jul 2026

1. Sanction listed: OFAC — RUSSIA-EO14024

   18 Jul 2026 · OFAC SDN

2. Alias recorded: MACHINE LEARNING LAB S.R.O.

   18 Jul 2026 · OFAC SDN

3. First recorded: MACHINE LEARNING LAB S.R.O.

   18 Jul 2026

No ownership test is on file for this company. The test reads who holds a company's shares from its shareholder register, and it is run each day for companies with a number in Estonia's, Latvia's or Poland's register. For a company anywhere else, no register read here states its shareholders, so the test cannot be run.

No corporate ownership records on file. This is an absence of evidence, not evidence of a simple structure — see the risk breakdown, where the ownership component is left unscored for exactly this reason.

Nothing to draw: this company has no recorded links to another company on file.

No address has been read for this company, so this was never checked — which is not the same as finding nothing.

Each reading is kept with the methodology version that produced it. A later reading is not a correction of an earlier one — it is a different question asked of more sources.

• Fact

  machine learning lab

  Legal name

  Evidence

  50172

  Read 13 Sept 2026

  Source

  OFAC SDN

  Primary source ↗

• Fact

  machine learning lab sro

  Legal name

  Evidence

  50172

  Read 28 Aug 2026

  Source

  OFAC SDN

  Primary source ↗

  Superseded 13 Sept 2026 — the source has since published a different value. Kept rather than overwritten, so the change itself stays readable.

• Fact

  CZ

  Jurisdiction

  Evidence

  50172

  Read 28 Aug 2026

  Source

  OFAC SDN

  Primary source ↗

• Fact

  OFAC — RUSSIA-EO14024

  Designation

  Evidence

  50172

  Read 28 Aug 2026 · asserted by 2 records

  Source

  OFAC SDN

  Primary source ↗

Records from these sources are matched into one company only on a shared identifier, or by a human reviewer — never on name similarity alone. How this works. Not legal or compliance advice: check the originals before acting on anything here.

TRUE

A source states it. Rendered as a fact, with the source attached.

FALSE

A source explicitly denies it. Rare, and never inferred from silence.

UNKNOWN

Checked, nothing found. Rendered as an em dash, never as 0 or "no".

NOT_ASSESSED

Never checked — nothing was read here. Excluded from coverage, and not counted as clean.

3 records on this company — identifiers, aliases, officers, listings, figures — each carry the source that wrote them and are shown beside their own entries. The fields above are different: the entities table has no source column, so what is shown is every source whose stored payload contains the value verbatim. That is strong evidence, not proof of which write landed it, and where nothing contains the value the row says so instead of naming a plausible source.