21 Aug 2026 AI accountability: Before or after the fact?
By WACC General Secretary Philip Lee
Reading the United Nations’ Preliminary Report of the Independent International Scientific Panel on AI (July 2026) – which offers an “evidence-based assessment of opportunities, risks and impacts of artificial intelligence” – reveals 16 references to accountability.
For example, “Transparency and accountability are key pillars for meaningful access to justice” (p. 38). No mechanisms are identified, with the excuse that, “Oversight is not yet operationalized as a measurable requirement with concrete expectations for intervention, reversibility and accountability as AI agents increasingly orchestrate other AI agents” (p. 23).
The risks, however, are all too apparent. “Recent years have seen rapid, and in some areas accelerating, progress in a range of AI capabilities. Significant investments in computing power, new AI methodologies, and specialized training data have led to sustained improvements in a wide range of AI capabilities… Limitations remain, such as in reliability, obtaining strong performance across human languages and cultures, interacting with physical systems, executing complex or multi-step projects and producing factual outputs” (p.8).
If AI were to be considered in the same light as a cure for cancer, it would be subject to all kinds of regulations, including lengthy testing in many different contexts and under many different circumstances. Yet AI is largely unregulated and its wide-ranging impacts unassessed.
The UN report is clear in its warning:
“Reliable methods for retaining control over highly autonomous AI systems are lacking. There are no scientific guarantees that AI agents will not violate instructions, and evidence is accumulating of cases where they already violate them. In laboratory settings, AI systems have been shown to violate their safety instructions to avoid being shut down. Similar behaviour may pose challenges to evaluation and oversight methods, as the ability of leading AI systems to recognize testing environments and produce misleading evaluation results that would favour their continued operation grows” (p.9).
Policymakers scramble to address this gap, while the general public avidly laps up the new technology’s capabilities. There is also a wide gap in public understanding of the potential for harm in a technology that has self-evident potential for good – a trait that governments and corporate interests are anxious to play up.
“Policymakers seeking to shape this governance face an evidence dilemma: they need evidence to make informed consequential governance decisions, but by the time the evidence exists, it might be too late to make them, as the evidence lags behind the pace of AI development. Dozens of distinct governance instruments that seek to embed ethics and human rights in AI systems are already in use across jurisdictions, but they are fragmented, are concentrated among a few corporations and rarely measure real-world effectiveness. Evaluation methods themselves are underdeveloped, and the institutions needed to provide independent capability and risk assessments remain embryonic” (p. 10).
Civil society’s response must be clear, urgent, and unequivocal.
- Awareness-building of the issues at stake must be a priority among people of all ages and all backgrounds, notably children and young adults.
- Tech companies must be reined in and held accountable with minimum general standards of safety and security.
- All new AI developments must be rigorously scrutinised for unintended consequences before being released and should be subject to a transparent and workable feedback mechanism so that flaws can be rectified.
These demands apply across the board but are especially relevant to the Global South.
As IT for Change warned in the run-up to the first Global Dialogue on AI Governance, “As things stand, AI governance is at a precipitous point, with the absence of political will to rein in the tech corporations. From mines to microchips, drudgery to data, genetic material to GenAI, synthetic data to missile strikes – we are witnesses to global AI value chains that brazenly disregard human dignity, social peace, and ecological well-being. The unprecedented geoeconomic power of the tech titans, fuelled by the US oligarchy, and the race to keep up with the innovation economy have eliminated considerations of global democratic accountability in AI governance.”
There are urgent calls for a global moratorium on the sale and use of AI systems that pose a high risk to human rights (such as remote biometric recognition, social scoring, spyware, and AI-driven autonomous weapons). The proposed UN Binding Treaty on Transnational Corporations aims to regulate global business activities under international human rights law and to hold multinational companies accountable for human rights abuses. It should be adopted without delay and appropriately future-proofed against the specific risks of harms and abuses in AI developments. Over ten years in the making, the treaty is stalled by government and corporate self-interest and intransigence, although there is talk of a final text being available by the end of 2027.
All elements of civil society, including the ecumenical movement and its many leaders worldwide, have a moral duty to speak up, to build awareness, and crucially to ensure that conversations about AI include the voices and experiences of those at the margins of digital societies and sustainable development. The communication justice mantra is clear – “Who benefits?” “Who knows?” “Who controls?”
The WACC delegation to Geneva Digital Week included project partners from Mexico and Tunisia taking part through our Communication for All Program (CAP), with support from Bread for the World – Germany.
Top image: Maria Ressa, co-chair of the Independent International Scientific Panel on Artificial Intelligence, presents the preliminary report of the 40-member body at the Global Dialogue on AI Governance on 6 July. Photo: ITU/D. Woldu on Flickr, CC BY-NC-SA 4.0.