By Dr. Gregory H Stanton, Founding President, Genocide Watch
Presentation to the IAGS/Educators Institute for Human Rights/Alliance Against Genocide
Conference on Teaching About Genocide
April 25, 2024
Dr. Gregory H Stanton, Founding President of Genocide Watch and Chair, Alliance Against Genocide
Using AI to Answer Hate Speech
A Proposal for an Artificial Intelligence Corporation to build AI to answer hate speech
By Dr. Gregory H. Stanton[i], Founding President, Genocide Watch
July 25, 2024
The problem of unanswered hate speech
In Myanmar, Facebook facilitated the spread of hate speech and incitement which directly led to the Rohingya genocide and the forced displacement of over a million refugees. Hatred in Myanmar against Rohingya was spread on Facebook and ignited genocide and forced deportation there. Hundreds of thousands of refugees remain in camps in Bangladesh.
Meta's WhatsApp is currently being used to spread hate speech and incitements to hate crimes in India. WhatsApp is being used to spread the conspiracy theory of “Love Jihad,” which portrays every mixed marriage as an insidious conspiracy that threatens the “purity” of Hinduism. It is an electrical charge intended to polarize Indian Hindus and Muslims against each other.
Kidnapped Yezidi women are still being sold by ISIS on Facebook today.
Rohingya and Yezidi plaintiffs and lawyers in Bangladesh and France are filing lawsuits to hold Meta accountable for its role in these genocides. Taking on a giant organisation like Meta will require enormous resources.
It may take such lawsuits to hold Meta, X, and other social media accountable for publishing hate speech on their platforms. Lawsuits on behalf of Rohingya and Yezidi victims may put Meta and other social media on notice that they will be held accountable for their complicity in genocide.
But lawsuits may not be the best way to convince Meta, X, and other social media to respond to hate speech. I am convinced that Mark Zuckerberg and Elon Musk and their employees do not really want to be complicit in genocide. They just want to maximize use of their platforms and increase their profits.
One of the failures of Meta, X, Google, YouTube, and other social media is that their owners are too often absolutists about free speech. They do not seem to understand that incitement to genocide (or to commit any crime) is not protected speech.
Meta has appointed a “Review Board” that always acts too slowly to take down hate speech and incitement in time to prevent it from “going viral.” The “Review Board” may be well intentioned. But it is too slow to prevent hate speech from creating cultures of genocide.
The solution is to instantly answer hate speech. Only AI is fast enough to do that.
The Genocide Watch Proposal
Genocide Watch proposes to start negotiations with the executives who control Artificial Intelligence corporations to jointly establish a non-profit corporation that will build, train, and reach agreements with the major social media platforms to include AI software that will instantly recognize and answer hate speech.
These AI programs will be trained to recognize hate speech and incitement in hundreds of languages. They would instantly answer hate speech, rather than just take it down. Only AI is fast enough to do that.
AI software that can recognize hate speech in hundreds of languages will cost billions of dollars. Founding and running this corporation will require the enormous resources of OpenAI, Anthropos, Google, X, Microsoft, Meta, and other Silicon Valley corporations. Only the giant IT companies have the money necessary to finance this project.
Convincing the owners of social media companies to incorporate such software into their platforms will require that they adopt this corporation and this project as their own. That is why it is necessary to create the corporation as a joint venture that draws on the strengths and expertise of the social media and AI companies themselves. The companies that will use the AI software will own it.
The project will be organized by Genocide Watch, the Alliance Against Genocide, and other partners in conjunction with the AI and social media companies that will own and invest in this non-profit corporation. The Alliance, AI, and social media companies and their associated law and venture capital firms have the legal skills and international organizational and financial networks to launch this corporation.
The Alliance Against Genocide is uniquely qualified to propose this project to the Silicon Valley corporations. The Alliance has 119 member organizations in thirty-one countries. Its member organizations have hundreds of staff in sixty more. Their staffs speak most of the world’s major languages.
The Alliance Against Genocide will work with the social media companies themselves to build this corporation. Social media and AI companies are the only organizations with the global capacity and financial resources to build and staff this AI corporation.
The corporation will be funded by investments, not by donations. Although it will be a non-profit, it will be run like an ordinary software company by business professionals. Its products will be adopted by AI and social media companies at the cost of producing them.
The AI programs to combat hate speech and incitement will require a vast reach because the AI programs will have to be trained in hundreds of languages.
The Alliance has established a partnership with the International Religious Freedom Roundtable which has members from many of the world’s religious denominations, including the Roman Catholic Church, which has churches in nearly every country. Churches, synagogues, mosques, and temples of the world religions have hundreds of thousands of priests, pastors, rabbis, imams, and staff in every country.
We will seek help from the Roman Catholic Church, the World Council of Churches, and Jewish, Muslim, Buddhist, Hindu, Chinese, and African traditional leaders, as well as from unaffiliated or non-religious academics, ethicists, social scientists, businesspeople, and AI experts.
The corporation will recruit the staff that will train the AI software to recognize and answer hate speech.
The AI programs that this corporation develops will have to be voluntarily incorporated into dozens of social media platforms. The AI corporation we envision will be a joint venture of the AI and social media companies brought together by Genocide Watch, the Alliance Against Genocide, and other partners..
Genocide Watch will propose founding this non-profit anti-genocide AI corporation with the creators and owners of the major social media platforms and AI software companies.
AI could make social media a guardian rather than a threat to mankind.
Dr. Gregory H Stanton
Founding President, Genocide Watch
[i] Dr. Gregory Stanton is a Yale Law educated corporate lawyer who practised international corporate law with a major U.S. law firm, Foley and Lardner of Milwaukee and Washington, DC. As a practising lawyer, Stanton specialized in establishing joint ventures across international legal systems. He then taught international business law and international human rights law at Washington and Lee University School of Law. During his six years in the U.S. State Department, he worked with the United Nations to establish international tribunals for Rwanda and Cambodia, for which he wrote the statutes and procedural rules. He proposed and lobbied for creation of the United Nations Office of the Special Adviser on the Prevention of Genocide. He was a Research Professor of Genocide Studies at George Mason University and the James Farmer Professor in Human Rights at the University of Mary Washington.