Cárdenas, Luján, Klobuchar Lead Colleagues Urging Tech CEOs to Combat Spanish-Language Disinformation
Lawmakers to CEOs: “Spanish-language moderation efforts are not keeping pace”
WASHINGTON, D.C. — U.S. Representative Tony Cárdenas (D-Calif.) and U.S. Senators Ben Ray Luján (D-N.M.) and Amy Klobuchar (D-Minn.) led 23 of their colleagues in letters to tech CEOs raising the alarm over the increasing rate of Spanish and other non-English language disinformation across Facebook, Twitter, YouTube and Nextdoor.
The lawmakers wrote, “We write to express our serious concerns regarding the increasing rate of Spanish and other non-English language disinformation across digital platforms, and the lack of transparency regarding efforts to limit the spread of this harmful content for all languages.”
“There is significant evidence that your Spanish-language moderation efforts are not keeping pace, with widespread accounts of viral content promoting human smuggling, vaccine hoaxes, and election misinformation,” the lawmakers continued. “Congress has a moral duty to ensure that all social media users have the same access to truthful and trustworthy content regardless of the language they speak at home or use to communicate online.”
The letter is cosigned by U.S. Senators Mazie K. Hirono (D-Hawaii), Catherine Cortez Masto (D-Nev.), Elizabeth Warren (D-Mass.), Bob Menendez (D-NJ), Alex Padilla (D-Calif.) Brian Schatz (D-Hawaii), Jack Reed (D-RI), Richard Blumenthal (D-Conn.), Mark Kelly (D-Ariz.), and Tim Kaine (D-Va.), and U.S. Representatives Kim Schrier (D-Wash.), Joaquin Castro (D-TX), Dina Titus (D-Nev.), Lori Trahan (D-Mass.), Jan Schakowsky (D-Ill.), Juan Vargas (D-Calif.), Debbie Dingell (D-Mich.), Ruben Gallego (D-Ariz.), Darren Soto (D-Fla.), Nydia Velásquez (D-NY), Raul Ruiz (D-Calif.), Raúl M. Grijalva (D-Ariz.) and Albio Sires (D-NJ.).
The letters to Facebook, Twitter, YouTube, and Nextdoor can be found HERE. The full text of the letter to Facebook is available below:
Dear Mr. Zuckerberg,
We write to express our serious concerns regarding the increasing rate of Spanish and other non-English language disinformation across Facebook’s platforms and your lack of transparency regarding efforts to limit the spread of this harmful content for all languages. We urge you to release specific and clear data demonstrating the resources you currently devote to protect non-English speakers from misinformation, disinformation, and illegal content on your platforms.
This letter follows previous letters sent to Facebook and Twitter, including a recent letter sent in April urging your platform to take action against the “Disinformation Dozen”. In addition, the letter calls for greater scrutiny on primary spreaders of vaccine disinformation and requests clear and concrete statistics regarding campaign efficacy and the number of full-time and contract level employees devoted to non-English content moderation. Unfortunately, both platforms failed to provide the requested information on investments made to non-English language content moderation through human review and algorithmic processes.
In your response, you failed to demonstrate the commensurate investment and the efficacy of those programs charged with limiting the spread of misinformation, disinformation, and illegal content. On the contrary, there is significant evidence that your Spanish-language moderation efforts are not keeping pace, with widespread accounts of viral content promoting human smuggling, vaccine hoaxes, and election misinformation.
Congress has a moral duty to ensure that all social media users have the same access to truthful and trustworthy content regardless of the language they speak at home or use to communicate online. Given the seriousness of this issue, we request additional information on what your company is currently doing to remove non-English misinformation, disinformation, and illegal content on Facebook, Instagram, and WhatsApp:
- What are the top five languages for content that users within the U.S. encounter on each of your platforms? Please provide an approximate percentage of users in the U.S. that encounter content within the given language on a weekly basis.
- How many of your employees and/or contractors provide content moderation in each of those languages? For each language, please break down further by including:
- Employment status of moderators (full-time, in-house, or contract-level), specifying the employees who work across multiple languages to avoid double counting;
- Country the employee is permanently stationed in; and
- Median hourly salary for the cohort broken down by employment status and country where the work is performed.
- Is all content reviewed in the original language of the post, or are some or all subject to automated translation before being reviewed?
- For your platform’s algorithmic processes that detect content for removal or additional review, do you have processes in place to ensure similar performance between all languages? If so, please describe. Please include in your response specific numbers that show investment or efficacy such as precision/recall metrics, the size of training data, or the number of full-time employees working exclusively on non-English algorithmic content detection.
We look forward to your prompt response, which we would request that you send us by Friday, August 13th, 2021.
###