Skip to content
Apart Research
Apart Research
Contact us
Sprint projectJun 22, 2026Hanoi

Shopee's Invisible Manager: Algorithmic Governance of ECommerce Sellers and the Regulatory Gap in Vietnam's AI Law

Võ Hồng Anh · Team Voh

Submitted to Global South AI Safety Hackathon. Sprint projects are early-stage work by participants, not Apart Research publications.

Read the report

Report: Shopee's Invisible Manager: Algorithmic Governance of ECommerce Sellers and the Regulatory Gap in Vietnam's AI Law

Open in a new tab (opens in new tab)
Share

Millions of small sellers on Shopee Vietnam have their income and platform access controlled by an opaque ranking and penalty algorithm — yet Vietnam's new AI Law (2026) classifies such systems as merely "medium-risk," requiring only basic transparency disclosures. This project investigates whether Shopee's algorithmic governance of sellers constitutes algorithmic management, and whether the current regulatory classification is adequate. Using policy document analysis, a seller survey (n=50–80), and comparative legal analysis against the EU AI Act, we identify a governance gap and propose concrete amendments to Vietnam's forthcoming implementing decrees.

Reviews

Judging this Sprint?

Review this project

Your public critique appears on this page without your name. Your private critique is not published; only the Apart team reads it. If you agree below, we share your review with grantmaking.ai (opens in new tab) and the Transformative AI Fund so strong projects can be funded.

Not shown on this page.

Shown on this page, without your name.

Only the Apart team reads this, and funders if you agree below.

Share my name publicly on grantmaking.ai *
Share my private critique with funders *

How much would this matter for AI safety if it worked? How innovative is it? For scores of 4-5: is this actually new to the field, or replicating recent work?

Scoring guide
  1. 1Negligible. No clear problem addressed, or no meaningful novelty.
  2. 2Limited. Addresses a real problem but with a generic or well-trodden approach. Incremental at best.
  3. 3Moderate. Clear problem with a reasonable approach; some novelty in framing or method beyond routine application of existing tools.
  4. 4Significant. Important problem with an original approach, or identifies a neglected problem area. A valuable contribution others could build on.
  5. 5Exceptional. Tackles a critical AI safety problem with a genuinely novel approach, or opens a new research direction. Clear theory of change. You'd be excited to share this with researchers in the area.

How sound are methodology, implementation, and findings?

Scoring guide
  1. 1Seriously flawed. Methodology broken, results uninterpretable, or implementation doesn't work.
  2. 2Weak. Approach has significant gaps: missing validation, flawed experimental design, or incomplete implementation.
  3. 3Competent. Technically solid given the short duration. Methodology makes sense, results are interpretable, limitations acknowledged, work builds toward clear conclusions.
  4. 4Strong. Thorough methodology with convincing validation. Results clearly support conclusions. Immediately useful for future work.
  5. 5Exceptional. Ambitious scope executed rigorously. Surprising findings, novel methods, or unusually robust validation.

How clearly are work, findings, and impact potential communicated?

Scoring guide
  1. 1Incomprehensible. Cannot determine what the project is actually claiming or doing.
  2. 2Hard to follow. Key information buried, missing, or diluted by excessive length. Significant effort to extract main points.
  3. 3Clear enough. Can understand the problem, approach, and results without undue effort. Core content clearly present: problem, method, findings, limitations.
  4. 4Well presented. Easy to follow, well-structured, appropriate level of detail. Target audience would get it quickly.
  5. 5Exceptionally clear. A pleasure to read. Complex ideas made accessible. Could serve as a model for how to present this type of work.

  1. The Vietnam's AI Law (Law No. 134/2025/QH15), effective March 1, 2026 mentioned:

    Article 9. Classification of risk levels of artificial intelligence systems

    Artificial intelligence systems shall be classified according to the following risk levels:

    a) A high-risk artificial intelligence system is a system that may cause significant harm to life, health, the lawful rights and interests of organizations and individuals, national interests, public interests, or national security;

    So it's outcomes-based (rights and interests of organizations and individuals), so if labor rights and interests are affected significantly, it can be seen as high-risk.

    Seller surveys need larger sample sizes and info on industry, income level,

  2. 4/2/4

    Criteria 1 - Impact Potential & Innovation: 4

    Criteria 2 - Execution Quality: 2

    Criteria 3 - Presentation & Clarity: 4

    Well-identified gap and used an algorithmic management framework.

    The design is sound, genuine triangulation where three methods independently test the central claim only challenge is no validated findings. The presentation is well-structured and easy to follow.

  3. Great work extending the algorithmic-management lens from gig work to e-commerce sellers. This is a genuinely neglected angle, and the regulatory hook (medium-risk classification for systems that function as high-risk employment-AI over sellers' livelihoods) is the strongest part. Proxy discrimination over millions of sellers' livelihoods with no recourse is squarely in AI-safety scope. My main hesitation is that the empirical backbone is a plan rather than a result, so the contribution rests on a design we can't yet evaluate. On the empirical side, the Facebook-recruited sample of 50–80 is exposed to the selection bias you flag, and the 60% opacity threshold reads as plausible but arbitrary. It would strengthen your argument to explain why 60% and not another. There's also an enforcement-realism question worth confronting: a high-risk label in a jurisdiction with limited capacity to audit a foreign platform's algorithm may be paper protection, so the analysis should address whether the institution that would act on this classification can actually function. The most valuable next step is to build the classification on that self-employment language and then draft the actual implementing-decree text you'd propose. Promising work on a real and well-framed problem.

    Read full reviewShow less

Cite this project

@misc{anh2026shopees,
  title = {{Shopee's Invisible Manager: Algorithmic Governance of ECommerce Sellers and the Regulatory Gap in Vietnam's AI Law}},
  author = {Võ Hồng Anh},
  year = {2026},
  month = jun,
  note = {Submitted to Global South AI Safety Hackathon, an Apart Research Sprint},
  howpublished = {\url{https://apartresearch.com/sprints/projects/shopees-invisible-manager-algorithmic-governance-of-ecommerce-sellers-and-the-regulatory-gap-in-vietnams-ai-law-qm75}},
  url = {https://apartresearch.com/sprints/projects/shopees-invisible-manager-algorithmic-governance-of-ecommerce-sellers-and-the-regulatory-gap-in-vietnams-ai-law-qm75}
}

More from Global South AI Safety Hackathon

All 216 entriesBrowse all projects

Build something like this at the next Sprint

AI Collusion Research Sprint · Oct 23 - 25, 2026