[Air-L] AI Summary - The World’s Growing Information Black Box: Inequity in Platform Research

[Joly MacFie]

https://www.techpolicy.press/the-worlds-growing-information-black-box-inequity-in-platform-research/

“The World’s Growing Information Black Box: Inequity in Platform Research”
by Rachelle Faust & Daniel Arnaudo (Nov 7 2025, TechPolicy.Press).
------------------------------
Context and overall thesis

The authors argue that we are entering—or have already entered—a phase in
which studying large digital platforms is becoming much harder, especially
for researchers and civil society organisations outside the US/Europe.
Whereas access to social-platform data used to be more open (what they call
the “golden era” of platform access), that era is over. With rising
restrictions, high cost of access and shrinking tool sets, the ability to
conduct independent, public interest research into how platforms work (and
how they can harm) is shrinking. In effect, the “information black box” is
growing.
They emphasise that this matters because platforms are central to public
discourse, election integrity, human rights, online harms, and yet the
richest data and tools are accessible mostly to well-resourced actors in
Europe/North America.
------------------------------
Evidence from the field

The article draws on a survey (28 responses) of “frontline defenders”
(fact-checkers, human-rights activists, civil society researchers) from
around the world. Key findings:

   -

   75 % of respondents said it has become more difficult to conduct
   public-interest research on platforms. (Tech Policy Press
   <https://www.techpolicy.press/the-worlds-growing-information-black-box-inequity-in-platform-research/>
   )
   -

   Over 85 % cited X (formerly Twitter) as a platform whose changes
   negatively impacted their work; 75 % cited Facebook; over 50 % cited
   Instagram. (Tech Policy Press
   <https://www.techpolicy.press/the-worlds-growing-information-black-box-inequity-in-platform-research/>
   )
   -

   Example: A civic-tech organisation in Iraq had previously had 10
   employees with comprehensive access via CrowdTangle; under the new system
   (Meta’s Content Library) only one employee was approved, greatly limiting
   capacity. (Tech Policy Press
   <https://www.techpolicy.press/the-worlds-growing-information-black-box-inequity-in-platform-research/>
   )
   -

   Smaller platforms (and languages, regions) are especially under-studied
   and under-resourced.
   -

   Research access isn’t only about tools/fees: the institutions of many
   civil-society actors (vs university academics) don’t have IRBs, may lack
   standing to apply, or may face heavy application burdens. For example,
   Meta’s Content Library (MCL) requires a formal proposal, IRB approval, etc.
   (Tech Policy Press
   <https://www.techpolicy.press/the-worlds-growing-information-black-box-inequity-in-platform-research/>
   )
   -

   Platforms’ data-access and “trusted partner” programmes have been scaled
   back, and prioritised for well-resourced countries/languages. (Tech
   Policy Press
   <https://www.techpolicy.press/the-worlds-growing-information-black-box-inequity-in-platform-research/>
   )
   -

   Regional/regulatory gaps: Many places outside the EU/US lack regulatory
   regimes (for example, rights to petition for platform data access) that
   permit researchers to access non-public platform data. (Tech Policy Press
   <https://www.techpolicy.press/the-worlds-growing-information-black-box-inequity-in-platform-research/>
   )
   -

   Although online harms (such as foreign interference via platforms,
   targeted abuse) remain substantial, the tools and data to study them are
   less accessible. (Tech Policy Press
   <https://www.techpolicy.press/the-worlds-growing-information-black-box-inequity-in-platform-research/>
   )

------------------------------
How platform practices deepen inequity

   -

   Platforms have increasingly shifted from open/sharing postures to a more
   closed, cost-centric model: sharing data is framed as risk, cost,
   potentially brand-damage. (Tech Policy Press
   <https://www.techpolicy.press/the-worlds-growing-information-black-box-inequity-in-platform-research/>
   )
   -

   The article points out a bias: most resources, content, and data-access
   support is in English and targeted at Europe/North America; this
   disadvantages many Global South researchers and languages. (Tech Policy
   Press
   <https://www.techpolicy.press/the-worlds-growing-information-black-box-inequity-in-platform-research/>
   )
   -

   Example: TikTok restricts its research API to “qualifying researchers”
   in the US/Europe — meaning many researchers in other geographies (even
   those facing serious platform-harm threats) cannot obtain access. (Tech
   Policy Press
   <https://www.techpolicy.press/the-worlds-growing-information-black-box-inequity-in-platform-research/>
   )
   -

   Platforms still monetise huge amounts of user data and train AI systems,
   while restricting independent access to that data for public-interest
   research. (Tech Policy Press
   <https://www.techpolicy.press/the-worlds-growing-information-black-box-inequity-in-platform-research/>
   )

------------------------------
Proposed framework for improved access

The authors outline a newly developed framework (by the Knight‑Georgetown
Institute) that aims to lower barriers to meaningful public-interest
research on platform data, while still respecting privacy. Key points:

   -

   The framework recognises the difference in needs between academic
   researchers (long-term, large-scale studies) and civil
   society/fact-checkers/human-rights actors (rapid, contextual, localised
   real-time harms). (Tech Policy Press
   <https://www.techpolicy.press/the-worlds-growing-information-black-box-inequity-in-platform-research/>
   )
   -

   It proposes “free access” to high-influence public platform data for
   public interest actors (to relieve financial burdens). (Tech Policy Press
   <https://www.techpolicy.press/the-worlds-growing-information-black-box-inequity-in-platform-research/>
   )
   -

   It defines “highly disseminated content” using two thresholds: an
   absolute threshold of 10,000 unique views/listens/downloads, and/or content
   in the top 2 % of weekly views in a given information environment
   (language/geography). (Tech Policy Press
   <https://www.techpolicy.press/the-worlds-growing-information-black-box-inequity-in-platform-research/>
   )
   -

   The framework is intended as a baseline — it is not perfect, not
   all-encompassing—but a foundation upon which platforms, regulators,
   research consortia can build. (Tech Policy Press
   <https://www.techpolicy.press/the-worlds-growing-information-black-box-inequity-in-platform-research/>
   )
   -

   The authors emphasise that future work must include voices from
   different geographies, languages, capacities so that the framework remains
   inclusive. (Tech Policy Press
   <https://www.techpolicy.press/the-worlds-growing-information-black-box-inequity-in-platform-research/>
   )

------------------------------
Significance & closing arguments

   -

   The article suggests that transparency in platform operations is vital
   to democratic health, human rights protection, and understanding online
   harms.
   -

   Without meaningful access to platform data — especially in the Global
   South or in non-English contexts — large swaths of digital life become
   “black boxes.”
   -

   The authors call on platforms, regulators and research actors to strive
   for greater transparency and inclusive access regimes so that we don’t end
   up with a world where only well-resourced actors in privileged geographies
   can study these powerful platforms.
   -

   In their framing: When everyday life and public debate increasingly
   happen online, enabling research into how platforms operate — in all
   contexts — is critical.

------------------------------



-- 
--------------------------------------
Joly MacFie  +12185659365
--------------------------------------
-