By Luc Steinberg, Head of Operations, Media Diversity Institute Global
The 20th Internet Governance Forum (IGF), convened by the United Nations, in Oslo this June, underscored the urgent need to defend media diversity as the debate around digital policy and the rise of artificial intelligence accelerate.
The IGF’s multistakeholder model brings together academics, NGOs, industry leaders, technical experts and policymakers. The UN’s ambitions for directing global digital policy include two landmark processes – the WSIS+20 review evaluating two decades of digital development promises, and last September’s adoption of the Global Digital Compact (GDC).
Yet the velocity of technological change threatens to outpace such consensus-building efforts to reign in Silicon Valley and has upped the stakes for journalism’s future.
New risks for journalism
Media revenue models are already in retreat due to tech disruption with decreasing trust and the stress of competing for advertising from Big Tech giants like Google and Meta, which together capture roughly 65% of global digital advertising spend. Recognising these challenges, the Dynamic Coalition on the Sustainability of Journalism and News Media (DC-Journalism) was established in 2019 under the Internet Governance Forum framework. The platform brings together press freedom advocates, media development organisations, and journalism sectors to address critical questions about the sustainability and viability of journalism. The coalition’s 2024-2025 report identifies a new existential threat to the industry: generative AI.
Companies like OpenAI, Anthropic, Google and X have all trained their AI models (LLMs) on journalistic content without compensation, transparency and very little accountability. Some publishers, like the New York Times, have brought legal cases against OpenAI, while others like Axel Springer have struck deals licensing their content. The report notes that the unauthorised scraping by AI companies often forces news organisations to put quality information behind paywalls “while AI-generated synthetic media proliferates”.
A recent Pew Research Center study found that Google users are already less likely to click on news articles if an AI summary appears at the top of their search results. The situation is one that further threatens the future of the open internet and the integrity of our information environment and consequently threatens the pluralism and diversity of media available to citizens.
Threats and opportunities in the Global South
While some big publishers can launch lawsuits and strike content deals, smaller publishers, especially those in the Global South, often have less resources and power at their disposal and there exists a knowledge and power asymmetry in how AI is developed and deployed between the Global North and Global South. Subscription models also tend to function better in places where people have more disposable income.
The report is not all doom and gloom though, and RNW note that some Global South publishers have used AI tools effectively to increase news distribution and boost audience engagement, which could “indirectly lead to gains for their business viability and sustainability”. Examples include adopting fundraising tools or creating more SEO-friendly headlines. In one example, Raseef22 (a partner on MDI Global’s MigraVoice project) uses AI for Arabic-to-Farsi translations. This does beg the question, however, if AI will begin to replace media workers or if small outlets can “AI their way” out of a systemic crisis as the DC-Journalism report pointedly asks.
Towards a solution
As the causes for the crisis in media viability are well-documented, and the negative effects of AI’s proliferation become known, the DC-Journalism report highlights measures that can be adopted by newsrooms and policymakers to help stem the tide, including:
- Anchor AI use in strong ethical standards (see the Paris Charter), retaining human oversight, transparency and clear labelling of AI-created content
- Reform copyright and competition rules giving publishers the power to consent and receive fair pay when their work is used on training AI models
- Pool resources via alliances (e,g the Journalism Cloud Alliance) to negotiate better rates for cloud and AI services as well as co-developing open-source tools
- Scale up AI training and media-literacy programs for journalists and youth to spot deepfakes
- Give journalism a seat at the table in national and international AI lawmaking.
- Expanding fact-checking collaborations and free, reliable deepfake detectors, especially around elections
While none of these actions alone would restore a once thriving independent media market, without them there will certainly be less room for independent media and that spells less space for diversity.
Conclusion
Multistakeholder dialogues move quite a bit slower than technological change, but the IGF and The Dynamic Coalition on Journalism (DC‐Journalism) show that it is possible to build consensus around what is urgently needed to fix our increasingly besieged media ecosystem.
By embedding ethical practices in the building of AI tools, building global alliances and updating regulatory frameworks we can safeguard a vibrant, inclusive media landscape for the decades ahead.
Read the full DC Journalism report here.
MDI is a member of GFMD
