UNESCO’S statement in response was a diplomatic shrug, underlining that they were prepared anyway for such an eventuality. After two previous similar actions by the United States, the Agency had worked to decrease reliance on US funding and increase contributions from elsewhere. One journalist thinks that the withdrawal will be a blow to UNESCO’s work, even while acknowledging UNESCO’s response.

Most know UNESCO as a body that preserves ancient temples and declares World Heritage Sites. However, the agency actively works to protect freedom of expression, counter hate speech, disinformation and misinformation, and promote ethical AI governance.

UNESCO has survived without US funding before. The United States left in 1984 and came back, left again in 2017 and then rejoined in 2023. But each withdrawal no doubt chips away at the Agency’s stability.

According to UNESCO’s statement, the US contributes only 8% of its budget. Therefore, leaving now is not catastrophic from a financial standpoint, but it is not nothing. Some of UNESCO’s programs may slow down if member States and philanthropies do not fill the void.

Politically, however, the US exit will create a vacuum to be filled by other actors, some whose ideas about rights and ethics may be at loggerheads with the international consensus.

If you are worried about algorithms amplifying hate speech, AI promoting misogyny, or governments weaponizing cybercrime laws to clamp down on democratic dissent, you should care about what happens to UNESCO. It is one of the few spaces where global standards on these issues are being set. It’s clear that global civil society must intensify actions to insure gains in these areas of rights.

The United States’ withdrawal from UNESCO is a pattern. This isn’t the first breakup, and it likely won’t be the last. The question is, who steps up next?

Photo: The UNESCO logo on the Hanoi Ceramic Mosaic Mural, the world’s longest mosaic wall, Hanoi, Vietnam. Credit: Claudine Van Massenhove

Meanwhile in New Zealand, MP Laura McClure proposed a bill to expand existing laws and criminalize the creation and sharing of non-consensual sexually explicit deepfakes. To support her proposal, she held up to her parliamentary colleagues an A3-size AI-generated nude image of herself created in less than five minutes on technology found through a simple Google search.

It emerges that some U.S. lawmakers who voted for the bill had not read the fine print.  Their hands are now tied for the next decade.

What to do about AI governance, what dimensions to regulate and how far, seem not to be clear to policy makers. Even while AI is the new frontier for criminal behaviour, it appears some legislators are still working out what the boundaries of digital communication should be.

Kenya is a case in point. This week, the government charged a software developer with “cybercrime” for creating a website to explain a controversial proposed finance bill in easy-to-understand language.

Meanwhile, in Brussels, legislators made important strides with the European Union’s AI Act, arguably the most comprehensively thought-out one so far. It identifies risk, outrightly bans some AI practices, demands transparency, and defines penalties for non-compliance, with some exemptions.

The eclectic nature of the global AI regulatory environment results in loopholes that can be easily exploited. Further, that regulation moves at the speed of bureaucracy while AI evolves at the speed of thought is an added complication.  

The stakes couldn’t be higher in this global experiment on who gets to control the future. Will it be governments protecting netizens or corporations chasing power and profit?

Photo: Shutterstock AI

The UN defines hate speech as “any kind of communication in speech, writing or behaviour, that attacks or uses pejorative or discriminatory language with reference to a person or a group on the basis of who they are, in other words, based on their religion, ethnicity, nationality, race, colour, descent, gender or other identity factor.”

While this definition is broader than “incitement to discrimination, hostility or violence”, which is prohibited under international human rights law, according to the UN it has three important attributes. It covers:

• Hate speech conveyed through any form of expression, including images, cartoons, memes, objects, gestures and symbols and it can be disseminated offline or online.
• Hate speech that is “discriminatory” (biased, bigoted or intolerant) or “pejorative” (prejudiced, contemptuous or demeaning) of an individual or group.
• Hate speech that denigrates real or perceived “identity factors” of an individual or a group, including: “religion, ethnicity, nationality, race, colour, descent, gender,” but also characteristics such as language, economic or social origin, disability, health status, or sexual orientation, among many others.

The identification of memes intended to misinform and/or incite violence depends on algorithm engineers. Those same people and/or their companies must, therefore, carry responsibility for whatever escapes their oversight. So far, algorithm engineers need to be qualified in computer science, mathematics, programming, and software engineering – but not in social ethics.

Commenting on the recent spate of riots in the United Kingdom, Nobel Prize winning journalist and critic Maria Ressa noted that everything about the violence that erupted on Southport’s streets, and then in towns across the country, was fuelled by wild rumours and anti-immigrant rhetoric on social media (The Guardian, 3 August 2024):

“You see this chain reaction in these alternative news channels, where disinformation can spread so quickly and can mobilise people to take to the streets – who are then prone to using violence because there’s this anger and these really deep emotions that are, of course, being amplified. And then, from these alternative outlets, it’s carried on to X or on to the mainstream social media platforms.”

In 2023, UNESCO put forward an Action Plan to regulate social media. In 2024, UN member states and governments will be asked to consider how to implement it. The Plan is based on seven key principles:

• The impact on human rights becomes the compass for all decision-making, at every stage and by every stakeholder.
• Independent, public regulators are set up everywhere in the world, with clearly defined roles and sufficient resources to carry out their mission.
• These independent regulators work in close coordination as part of a wider network, to prevent digital companies from taking advantage of disparities between national regulations.
• Content moderation is feasible and effective at scale, in all regions and in all languages.
• Accountability and transparency are established in these platforms’ algorithms, which are too often geared towards maximizing engagement at the cost of reliable information.
• Platforms take more initiative to educate and train users to think critically.
• Regulators and platforms take stronger measures during particularly sensitive moments like elections and crises.

TikTok and Telegram were prime collaborators in instigating the UK riots. Effective regulation is vital and urgent. But it is clear that without long-term cooperation between governments, regulatory authorities, civil society and the platforms themselves, social media will continue to be used by racist thugs to disrupt and hurt society.

Photo: A protestor prepares to throw a brick towards riot police in Southport, Merseyside, UK, on 30 July 2024.
Credit: Ian Hamlett/Shutterstock

Apple has been fined €1.8bn by the European Commission for breaking competition laws over music streaming. The EU Competition Commissioner said that Apple had restricted “developers from informing consumers about alternative, cheaper music services available outside of the Apple ecosystem.”

Although Apple says it will appeal, the judgment seems part of the budding effort by some governments to address the huge power of the tech giants such as the parent companies of Google and Facebook, Microsoft, Apple, Amazon, and – rising rapidly due to its involvement in AI – Nvidia.

The EU’s Digital Services Act and Digital Markets Act are perhaps the most comprehensive to date in providing a legal framework to protect users, protect fundamental rights, and ensure fair use. The US has strengthened antitrust enforcement in the digital technology arena, which has rare bipartisan support.

Timothy Wu, in The Master Switch: The Rise and Fall of Information Empires published in 2012, charted a cycle of invention, growth, concentration, regulation, and competition over communication technologies such as telephone, radio, and the internet.

“History shows a typical progression of information technologies,” he writes, “from somebody’s hobby to somebody’s industry; from jury-rigged contraption to slick production marvel; from a freely accessible channel to one strictly controlled by a single corporation or cartel — from open to closed system.”

An experimental and open technology will inevitably become more concentrated, he argues, because all forms of communication require a network of users. And when more people join one network, the people they are connected to want to join as well. And the larger the network, the more opportunity for profit.

Communication companies then try to grow larger and larger, buying up smaller entities, until they become monopolies that set the financial terms and control the message.

With the internet and digital communication technology, the potential of the profit – and control – is even more obvious as it brings all print, radio, television, photograph, and telephone into one essential platform.

History, though, has also shown that while sometimes governments allow companies to grow bigger, they are also the one entity that decides when big is too big and can wield the power to break up a corporation. The speed of the digital transformation has meant the financial size of Big Tech companies is greater than most countries, so the US and EU regulatory actions are important first steps in starting to draw the line on the concentration of wealth and control. 

Consumers and rights advocates are essential actors in holding both companies and governments to account. From the products we buy and use to the votes we cast and the campaigns we engage in, we all have to speak and act for equality, equity, transparency, participation, and accountability.

It is far too early to tell for sure, but between the recent regulations and consumer advocacy, we may be at the beginning of a turning point for Big Tech.

Illustration: Shutterstock

There is however a spectrum of reasons for opposition to such regulation, from concerns about potential for abuse by authoritarian States to global big tech’s interest to operate unfettered on their own terms.

Case study: Canada

The scenario playing out in Canada is instructive.

Over three months ago, Meta began taking steps to end news access in Canada on its social media platforms, setting off a call by journalists in Quebec for audiences to take a 24-hour break from Facebook and Instagram and to subscribe to a local media outlet.

Since Meta’s blackout started, digital media expert Alfred Hermida estimates that between 20% to 30% of audiences of local news outlets have disappeared.

In contention is Canada’s Online News Act, set to enter into effect fully no later than December 19, 2023. According to the government, the Act’s objective is to “ensure that dominant platforms compensate news businesses when their content is made available on their services.”

A raft of regulations is proposed, including requiring platforms to conduct an open call process to solicit news business, to provide safeguards to protect journalistic independence, and to support the participation of collectives of small independent news outlets.

Mutually beneficial regulation?

I asked ChatGPT what it would take for social media corporations to pay publishers of news content shared on their platforms. The drivers it cited included regulatory intervention, industry agreements, fair compensation models, increased public and media pressure, and global collaboration.

“Ultimately, finding a sustainable and mutually beneficial model will likely require collaboration, open dialogue, and a willingness to address the concerns and interests of all stakeholders involved,” the generative AI tool told me.

Introduction of similar legislation two years ago in Australia produced a different outcome with significant rewards eventually for various media groups, according to Hermida, who is a journalism professor in western Canada. The difference came from the Australian government’s approach of not naming specific digital companies in the law’s formulation.

Not “if” but “when”

There’s a clear interest in regulation of digital corporations. Two waves of a UK public opinion poll revealed a decline in the proportion of respondents who felt that the digital technology sector is sufficiently regulated.

Almost three out of four persons in another survey covering 36 countries stated that technology giants should see their powers limited.

The question is not “if” but “when.” The news blackout in Canada by big tech aggregator platforms is likely here to stay, that is, until a solution emerges.

To paraphrase Theresa Blackburn, an independent news outlet owner in eastern Canada, the onus is on the platforms themselves that, in fact, already bear responsibility for the demise of numerous small news organisations.

Freedom on the Net 2023: The Repressive Power of Artificial Intelligence assesses internet freedom in 70 countries, accounting for almost 90 percent of global internet users.

Their headline findings are depressing but perhaps not surprising:

1. Global internet freedom declined for the 13^th consecutive year.
2. Attacks on free expression grew more common around the world.
3. Generative artificial intelligence (AI) threatens to supercharge online disinformation campaigns.
4. AI has allowed governments to enhance and refine their online censorship.

What provides constructive hope, however, are the report’s assessment of regulatory models (or lack thereof) and suggestion of a road map, largely built on recent steps taken by the European Union (EU) and United States.

The EU’s General Data Protection Regulations are a key foundation for the draft Artificial Intelligence Act, which “would tailor obligations based on the level of risk associated with particular technologies.”

The US effort, “The Blueprint for and AI Bill of Rights”, focuses on principles to AI design, use, and deployment but to date is dependent on voluntary commitments by companies.

The report notes that business decisions over the research period alone, notably by new owner of X (formerly Twitter), cast doubt on the effectiveness of self-regulation and make the case for responsible oversight more acute.

The report’s data and recommendations highlight the continued need for government, companies, and civil society to work together with a rights-based approach that first and foremost protects the freedom of expression and access to information. Regulation should be based on human rights, transparency, and independent oversight.

The recommendations also address the need to defend “information integrity” but admit that this requires a long-term solution that prioritizes independent media and educated communities:

A whole-of-society approach to fostering a diverse and reliable information space entails supporting independent online media and empowering ordinary people with the tools they need to identify false or misleading information. Civic education initiatives and digital literacy training can help people navigate complex media environments. Governments should also allocate funding to develop detection tools for AI-generated content, which will only become more important as these tools grow more sophisticated and more widely used. Finally, democracies should scale up efforts to support independent online media through financial assistance and innovative financing models, technical support, and professional development support.

Reading through the report data and recommendation, though, one wonders what would ever convince authoritarian governments (which seem to be increasing in number) and globally powerful tech companies to cede their power to control. And certainly, viewing the global reality is overwhelming.

But the power of democratic advocates and activists can be seen at local and national levels – and these are the real building blocks of change. As the report itself notes with several examples, “Digital activism and civil society advocacy drove real-world improvements for human rights during the coverage period.”

These are the positive cases that we need to work together to multiply.

Photo: Dmitry Demidovich

Human rights activists are concerned that such data can be used to discriminate against migrants, refugees, and on racial grounds.

The “Report of the Special Rapporteur on contemporary forms of racism, racial discrimination, xenophobia and related intolerance” (United Nations, November 2020) is categorical:

“Governments and UN agencies are developing and using emerging digital technologies in ways that are uniquely experimental, dangerous, and discriminatory in the border and immigration enforcement context. By so doing, they are subjecting refugees, migrants, stateless persons and others to human rights violations, and extracting large quantities of data from them on exploitative terms that strip these groups of fundamental human agency and dignity.”

The focus of the report is the prevalence of digital technologies in immigration and border enforcement. Some commentators are referring to the rise of “digital borders” – meaning those whose infrastructure and processes increasingly rely on machine learning, automated algorithmic decision-making systems, predictive analytics and related digital technologies.

The UN Special Rapporteur, Tendayi Achiume, was interviewed by Katy Fallon in “UN warns of impact of smart borders on refugees: ‘Data collection isn’t apolitical’” (The Guardian, 11 November 2020). She said, “One of the key messages of the report is that we have to be paying very close attention to the disparate impact of this technology and not just assuming that because it’s technology, it’s going to be fair or be neutral or objective in some way.”

The article noted that Covid-19 has accelerated “biosurveillance” – tracking people’s movements and health. Citing the “Covi-Pass”, a health passport developed by Mastercard and Gavi, a private-public health alliance, being tried out in several African countries, the UN Report criticised the potential impact of such passports on freedom of movement especially for refugees.

In response, the Special Rapporteur has proposed a structural and intersectional human rights law approach to the design and use of emerging digital technologies. It emphasises the scope of legally prohibited racial discrimination in their design and use; current obligations to prevent and combat racial discrimination; and obligations to provide effective remedies for racial discrimination.

It’s clear that civil society groups need to monitor the ways digital technologies are being applied in humanitarian crises and elsewhere in order to challenge discrimination and injustice.

Photo: Alexander Kirch/Shutterstock

Under new regulations announced by the Tanzania Communications Regulatory Authority, which came into force on 10 August 2020, local media must now seek government permission to broadcast foreign content. They will be responsible for any perceived “offence” contained in that content.

John Magufuli, President of Tanzania. Photo: Wikimedia Commons

Worse still, the holder of a licence is not allowed to visit, or do business with, foreign nationals regarding the broadcast of content without being accompanied by a government official or an employee of the regulatory authority.

Since taking office in 2015, the administration of President John Magufuli has expanded its definition of “offence” to include almost anything contrary to his personal views on politics, the economy, and Covid-19. The Tanzanian president has publicly disparaged social distancing and wearing facemasks as well as promoted a herbal tea from Madagascar as a cure.

The Tanzanian public sphere is shrinking. According to Reporters Without Borders, three web TVs and a leading daily newspaper were the victims of arbitrary bans in 2019. By closing 15 media outlets over the past four years and by constantly threatening to withhold state advertising from privately owned media, the government has created a climate of fear in which self-censorship is growing.

In addition, online information has also been restricted since the adoption of a draconian law under which websites and blogs have to pay exorbitant fees to register and get accreditation.

Tanzania is not alone in trying to control social media. A law passed by the Turkish Parliament on 29 July 2020 introduced sweeping new powers. It obliged social media platforms with more than one million daily users to appoint a local representative in Turkey. Activists fear this will enable the government to carry out even more surveillance and censorship.

The new legislation forces platforms to store user data in Turkey and to maintain offices in the country. When the authorities find a tweet problematic, for example, they will force the Turkish office of Twitter to remove it. If the company refuses, the government will hit it with heavy fines. The same for Google’s YouTube videos.

Freedom of opinion and expression includes the freedom to seek, receive, and impart information and ideas without hindrance through any media and regardless of frontiers. Governments ignore such rights at their peril.

Photo above: wk1003mike/Shutterstock