Meta’s first human rights report, released last month, wasn’t just mediocre. It was a reminder of the dangers involved in using human rights language to whitewash bad corporate behavior. Many in civil society were intrigued, if not exactly hopeful, when Meta created a human rights team. We hoped it meant the company would start to take accountability, and provide an easy path for us to engage with the right people to implement meaningful changes based on the feedback we provided. We were also nervous that Meta would use its creation of a human rights team to whitewash bad behavior and sidetrack advocates seeking to engage with decision-makers who can make policy, enforcement, and engineering changes. Unfortunately, our fears appear to have been justified.
That’s why Mnemonic joined with 14 other civil society organizations and individuals to call on Meta to immediately release its human rights impact assessment in India, and its independent review of its content moderation activities in Palestine. Our letter has four clear demands:
- Publicly release, without further delay, the full and unedited report for Palestine that includes all of BSR’s findings and recommendations.
- Release the India human rights impact assessment with all of Foley Hoag’s findings and recommendations, and explain associated security concerns in greater detail to Indian civil society.
- For both Palestine and India, and for all HRIAs and independent reviews going forward, give the civil society organizations that contribute to the reports the ability to review all findings and recommendations, even those that cannot be made public.
- Going forward, when Meta contracts with companies or law firms to conduct independent reviews, HRIAs, or other forms of due diligence they must make public the terms of Meta’s engagement with the third party conducting the audit, in particular who has the authority to review and release the report.
As noted in the letter, we felt it important to act because the report sets a bad precedent for Meta’s use of independent review in any form, whether a human rights impact assessment, civil rights audit, or simply a review. Increasingly, civil society and legislators are calling for various forms of third-party review. The report also mentions such reviews repeatedly. We want Meta to understand that just doing the review isn’t enough. They have to be meaningful, and they have to be released. The whole point of such transparency is, ultimately, accountability.
Instead of taking any accountability, the report makes it clear that Meta wants credit for every single public effort it has made to support human rights, but it doesn’t want to acknowledge any of its failures, nor all the work civil society has done to address those failures. The report was a glowing description of every one of Meta’s policies and processes related to human rights. What it was missing is any admission of where the company truly still has work to do.
Worse, the report downplays the herculean efforts of civil society that have been required for nearly every positive change that has happened in the last decade at Meta. In fact, it claims that the Oversight Board, “via its case decisions and policy recommendations, is defining what a human rights-based approach to content means in practice, to a level of detail, practical application and intellectual rigor, that did not previously exist.” The Oversight Board has made great decisions, which incorporate many of the suggestions in our comments on individual cases, as well as many of the recommendations civil society had already made for years before the Board existed. Intellectual rigor is subjective, but one can simply look at the many open letters civil society, including Mnemonic, has sent to Meta over the years to understand that detail and practical application have never been lacking.
There are a number of areas in the report where Meta could have simply noted that it was still figuring out the best path forward, but instead glossed over major problems. For example, the report comments on the salient human rights risk to ” Right to Life, Liberty and Security of Person.” The report ignores the connection between misinformation and offline harm- one disturbingly on display in India as content connected to the dangerous and Islamophobic concepts of “Love Jihad” and “Corona Jihad” was left up for weeks or even indefinitely after it was reported to the company.
Another major oversight is the report’s summary of the company’s response to COVID, and the lasting impacts of that response. There is no mention of the increased use of automation at the beginning of the COVID 19 pandemic, which resulted in increased takedowns. There is also no update on the decreased availability of reviewers, nor anything about what is actually happening with appeals now, despite the fact that Facebook’s help center still notes the company has “fewer reviewers available right now because of the coronavirus (COVID-19).”
There’s nothing wrong with Meta writing about areas where it has improved or changes that it has made. The problem with this report is that it does so to the exclusion of any meaningful insight or accountability. That’s not especially surprising, since the company is writing about itself, and it has to cater to shareholders and lawmakers, not just civil society. That’s why third-party reports are so important. And that’s why civil society has reiterated our call for Meta to publish the full India HRIA and Palestine Independent Review immediately.
Read the full letter here