TECHNOLOGY: INNOVATION WITHOUT ACCOUNTABILITY
- Tech oligarchs have accumulated vast power and are using it irresponsibly. Many have aligned themselves with authoritarian, nationalist and populist leaders, including by forging links with the Trump administration and publicly supporting extremist movements. Convergence between military, political and technological elites is enabling conflict atrocities and threatening civic space and the rights of excluded groups.
- Technology companies are rapidly rolling out generative AI without adequate human rights safeguards while regulations haven’t kept pace. AI is amplifying electoral disinformation, powering military kill lists, driving surveillance technology that’s used against civil society and enabling deepfake pornography that overwhelmingly targets women and girls. States are failing to regulate out of fear of missing out on economic and military advantage.
- Civil society is leading the campaign to make technology more responsible, providing fact-checking initiatives, litigating against tech companies over hate speech and its amplification in conflict settings, exposing corporate complicity in human rights abuses and demanding global standards for technology, including AI.
Technology companies are rapidly rolling out generative AI as they seek to get ahead of a potential financial bubble burst. Under intense competitive pressure, tech companies have de-emphasised ethical commitments, unleashing products regardless of potential human rights impacts. Tech leaders are also increasingly showing they can’t be trusted by aligning themselves with politicians who attack rights. Companies are exploiting governance gaps, since technologies develop faster than regulations, and states, increasingly captured by or deferential to tech interests, are failing to act.
Civil society is encouraging responsible use of new technologies while exploring their activist uses. Social media plays a key role in mobilising Gen Z-led resistance to unaccountable political and economic power. In repressive contexts, diaspora activists use social media to work with those at home. As funding sources collapse and civic space restrictions deepen, civil society groups are using AI tools in campaigning, research and grant applications.
Civil society is using tech tools to pursue accountability for rights violations. Investigators in Ukraine use AI and open-source tools to identify and collect war crimes evidence. Mexican civil society has created a platform to help families identify disappeared people from recovered belongings. Young activists from Myanmar have developed a platform to help people organise securely in conflict zones.
At the same time, states and corporations use digital tools to restrict and repress civil society. Social media companies have shown themselves willing to de-prioritise and shadow-ban activist content and apply opaque moderation policies, while failing to deal with disinformation, hate speech and threats. Authoritarian states impose internet shutdowns, limit social media access and criminalise activists and journalists for online expression.
Journalists protest against ‘fake news’ legislation used to criminalise opinion and silence dissent in Lahore, Pakistan, on 31 January 2025. Photo by Arif Ali/AFP
Governments in Cameroon, Tanzania and Uganda imposed internet shutdowns around non-competitive elections. Iranian authorities shut down internet access during their lethal repression of anti-government protests. Nepal’s government blocked 26 social media platforms, quickly backtracking in the face of Gen Z-led protests. Faced with Gen Z-led protests, Turkish authorities severely limited bandwidth for key social media platforms. Indian authorities ordered the blocking of around 10,000 Twitter/X accounts amid conflict with Pakistan over Kashmir.
States use cybercrime and so-called fake news laws to criminalise online dissent. Mali’s military junta uses cybercrime laws against activists, journalists and opposition politicians, deploying catch-all offences such as ‘undermining state credibility’. In repressive Belarus, the government detains people for following Telegram channels it deems extremist. In Indonesia, eight activists face long sentences for supporting Gen Z-led protests on social media.
Kenyan activist Albert Omondi Ojwang was killed in custody after his arrest for criticising a police official on social media. Turkish authorities returned two exiled bloggers to Turkmenistan on false terrorism charges; their whereabouts remain unknown.
Tech companies often align with repressive states, complying with requests to take down accounts and provide user data. YouTube, for example, has given in to Israel’s demands to block Al Jazeera livestreams.
Disinformation thrives online, and regressive voices are often the loudest. Algorithms promote sensationalist content, while civil society’s more nuanced messaging struggles to compete.
Populists and nationalists have followed Donald Trump’s lead, exploiting social media to build their brands. Argentina’s President Javier Milei bypassed traditional media to cultivate an anti-establishment persona through meme-heavy content and AI-generated Instagram and TikTok imagery. Japan’s far-right Sanseitō party far outstrips the ruling party’s YouTube numbers, which helps it recruit young voters. The far-right Alternative for Germany party has a strong TikTok presence. Burkina Faso’s military ruler Ibrahim Traoré has won a following through disinformation that plays up his charisma and minimises rights violations.
AfD federal chairwoman Alice Weidel is interviewed for the party’s YouTube channel during the national party conference in Essen, Germany, on 29 June 2024. Photo by Bernd von Jutrczenka/DPA via AFP
Politicians use social media reach to attack civil society. Slovakia’s populist prime minister Robert Fico boosted an online smear campaign against a reproductive rights activist. A far-right Portuguese politician published the names of migrant students, accusing them of taking spaces from Portuguese-born children and calling activists who support them extremists.
Amid an online pandemic of misogyny, women activists are particularly vulnerable to gendered attacks, including doxxing, online stalking and deepfake porn, supercharged by AI. The danger is that women disengage and leave the space free for those who spread hate. Online vilification drives violence, as LGBTQI+ Tunisians can attest. Online climate denial is rife, with Spanish authorities warning about growing hate speech against climate researchers.
Tech companies have intentionally boosted right-wing disinformation. In Poland’s presidential election, TikTok’s algorithm served new users far-right content twice as often as centrist or left-wing content. In Ethiopia, Facebook stands accused of algorithmically boosting hateful content during conflict.
Alongside Poland, elections in the Czech Republic, Moldova and Romania were awash with pro-Russia disinformation. Moldova saw Russia’s most intense interference attempt, with people paid to covertly spread propaganda on Facebook and TikTok. In Romania, the Constitutional Court controversially annulled an election following evidence of a large-scale TikTok manipulation campaign.
Disinformation campaigns often target civil society. In Thailand, leaked documents exposed the existence of a secret government team running online smear campaigns against organisations including Amnesty International. Syria’s White Helmets, a volunteer rescue group, routinely faces disinformation branding it as a terrorist network.
Tech companies are enabling states to put civil society activists and organisations under surveillance. One method is spyware: Togo’s authoritarian government has deployed it against journalists. In Italy, the government hacked the phones of activists and journalists, including people working on migrants’ rights. In Serbia it was used against a student activist involved in Gen Z-led protests.
The Metropolitan Police deploys a live facial recognition system at Oxford Circus in London, England, on 13 May 2025. Photo by Leon Neal/Getty Images
The most widely used spyware, Pegasus, and Graphite, the spyware used in Italy, are produced by Israeli companies. Israel is a world leader in these hostile technologies. Its tech companies are closely entwined with its military forces, and sell only to states, although complex vendor networks obscure where spyware ends up. The Israeli government shields its tech companies; its refusal to cooperate forced a Spanish court to drop a spyware investigation.
Protesters are threatened by the growing use of facial recognition technology. AI enables live facial recognition without human oversight. Introduced on security grounds, this technology normalises intrusive surveillance and pre-emptive policing. Authoritarian states have pioneered it and China exports it, including to Afghanistan. Russia uses it to pre-emptively intercept people who may protest against its war on Ukraine. Turkey used it to identify participants in the 2025 democracy protests. Hungary’s new law banning Pride events is enforced through facial recognition.
AI also enables systematic surveillance of social media activity. The US government uses AI to scan accounts of student visa holders, looking for expressions of sympathy for Hamas to justify deportations. It uses automated vehicle licence plate readers and doorbell camera technologies to identify migrants and protesters, while Iran’s theocracy uses licence plate readers to enforce restrictions on women. Israeli police now have authority to control cameras and microphones in private devices.
Tech companies are working openly with authoritarian regimes. When tech companies and leaders donated millions to Trump’s inauguration fund, they abandoned any pretence of social responsibility. Companies such as Amazon, Google and Meta quickly dropped or downplayed diversity, equity and inclusion initiatives, while Meta announced the end of its independent fact-checking programme in the USA.
Trump has repaid tech leaders generously, opposing regulations and handing them greater power. Several have become his advisers, tightly connecting defence, political and tech elites. The Trump administration is shaping the tech landscape, forcing the sale of TikTok’s US arm to a consortium including Oracle, owned by ally Larry Ellison. This threatens to turn the platform into a surveillance tool against people targeted by the Immigration and Customs Enforcement agency and those who protest against state violence.
Tech executives Mark Zuckerberg, Jeff Bezos, Sundar Pichai and Elon Musk attend the 60th presidential inauguration at the US Capitol Rotunda in Washington, USA, on 20 January 2025. Photo by Julia Demaree Nikhinson/Pool via Reuters/Gallo Images
Ideological alignment is evident. Trump allies include Peter Thiel, the Antichrist-obsessed extreme right-winger who owns data broker Palantir. The company provides technology that supports Israel’s genocide and US deportations. Twitter/X owner Elon Musk worked with the Trump administration in its chaotic cost-cutting operation. In 2025, he addressed a UK far-right rally and endorsed the Alternative for Germany party, while his AI chatbot Grok has repeatedly generated claims that ‘white genocide’ is unfolding in South Africa.
In 2026, Musk was embroiled in another controversy when the platform enabled users to post AI-generated nude pictures of women. This so-called nudification dehumanised women online; some children were also targeted. When the UK introduced plans to outlaw the creation and supply of nudification tools, Musk was forced to block it in countries where it’s illegal, but not before offering to limit access to subscribers, as though misogyny were a privilege that could be purchased.
Time and again, women and girls are the targets, and in cases like this, problems stem less from regulatory omission than from deliberate political choices. Toxic masculinity, the far right and the unaccountable power of tech oligarchs are interconnected problems.
The ubiquity of technology provided by US firms is also fuelling rising concerns about data security and the potential for the US government to use denial of access as a coercive tool. This has led to the French government’s decision to pursue digital independence by replacing US technology. France can expect retaliatory threats, while less powerful states won’t be able to make the choice.
Tech giants are also spearheading a literal AI arms race, increasingly partnering with military forces, even if it means rewriting mission statements that used to rule out military cooperation.
Activists from the Campaign to Stop Killer Robots stage a protest at the Brandenburg Gate in Berlin, Germany, on 21 March 2019. Photo by Annegret Hilse/Reuters via Gallo Images
Israeli forces use algorithmic systems to generate kill lists and track Hamas personnel so they can bomb them at home with their families. Several states now possess autonomous weapon systems – killer robots – that can select and strike targets without human intervention. The UK government recently signed a deal with Palantir to develop an AI-assisted ‘kill chain’. Automated systems lack conscience, empathy and an awareness of consequences. Because they rely on pattern recognition and probability calculations, and their algorithms are trained on biased data, they make mistakes.
The technology is advancing quickly and may soon become too entrenched to backtrack. Talks under an existing treaty, the United Nations (UN) Convention on Certain Conventional Weapons, have been slow, with major powers blocking progress. In December 2024, a UN General Assembly resolution established a new process to develop global regulations, with talks held in 2025.
The Campaign to Stop Killer Robots is playing a leading role. Most states support its proposed two-tiered approach, which would ban the most dangerous autonomous weapons and regulate the rest.
Civil society is also working to hold tech companies accountable for their role in Israel’s genocide. A Dutch group revealed that a Microsoft data centre in the Netherlands stores data for Israeli military intelligence, which uses cloud-based services for systematic surveillance of Palestinians. Civil society is urging employees to protest and is working to spread awareness about tech companies’ complicity in atrocities.
The rapid spread of generative AI has exposed a fragmented regulatory landscape. Neither states nor corporations can be trusted to develop regulations alone. Given the transnational nature of technology, civil society is calling for global governance standards rooted in international human rights law. Standards should be developed through consultative processes, and civil society should play a central role in oversight.
There’s been some regulatory progress at the European Union (EU) level. The European Media Freedom Act protects journalists against spyware, although states can abuse exceptions. In 2024, the EU adopted the AI Act, to be rolled out over several years. In 2025, it opened a voluntary code of AI practice for tech companies to sign.
The AI Act takes a tiered approach based on risk, banning high-risk applications and imposing transparency requirements on lower-risk ones. Civil society influenced its development but has major concerns, including over AI applications in defence, law enforcement and migration management.
US President Donald Trump gestures after signing an executive order at the ‘Winning the AI Race’ Summit in Washington, DC, USA on 23 July 2025. Photo by Andrew Caballero-Reynolds/AFP
But tech leaders are furious about constraints on their power, and they have Trump’s backing. Under their pressure, in November the EU proposed delaying the Act’s rollout until December 2027. The Trump administration has also spoken out against the EU’s General Data Protection Regulation, which governs personal data management. EU states may end up watering down AI and other regulations to avoid tariffs.
China has proposed its own rules, strongly centred on economic development, security and state sovereignty rather than human rights. The USA’s 2025 AI Action Plan reflects the influence of tech leaders, prioritising deregulation and asserting the goal of US global AI dominance. At the AI Action Summit in Paris in February 2025, the USA and UK refused to sign a declaration supporting ethical, inclusive and open AI governance. The latest summit, in India in February 2026, failed to make any progress on human rights safeguards.
In August 2025, the UN took a step forward with a resolution that created a panel of independent scientific experts and a Global Dialogue on AI Governance. Achieved after extensive negotiations, the resolution sought a compromise between the approaches of China, the EU and the USA. More difficult negotiations lie ahead, and it’s hard to be optimistic given that powerful states are increasingly undermining the UN and tech companies continue to use their power to weaken regulatory efforts.
In September, Nobel Peace Prize laureate Maria Ressa launched a Global Call for AI Red Lines, backed by hundreds of public figures and organisations. It urges governments to establish clear boundaries on surveillance and weapons systems, and calls for clear and enforceable rules, robust compliance mechanisms and a strong enforcement body. It offers an example of how civil society is taking the lead in shaping safeguards and standards.
A troubling development has, however, given repressive states a new weapon: the UN Convention against Cybercrime. Adopted in December 2024 and opened for signature in October, 74 states have already signed the treaty, suggesting it will pass the ratification threshold and enter into force. Civil society has consistently criticised the convention, which was proposed by Russia and backed by authoritarian states.
Although civil society engaged with the process, the treaty lacks adequate rights safeguards, and its broad scope gives states ample room for abuse. It may help combat harmful cybercrimes such as the non-consensual sharing of intimate images, but could intensify the use of cybercrime laws to suppress dissent or facilitate the extradition of dissidents in exile. How states incorporate the treaty’s provisions into their criminal justice systems will be crucial. Civil society needs a strong voice in reviewing implementation.
Regulation can have unintended consequences. When the EU adopted its Transparency and Targeting of Political Advertising regulation in 2024, it aimed to safeguard democratic processes from disinformation through stricter rules on political ad targeting. But in July, Meta claimed it couldn’t distinguish between political and non-political ads and announced it would no longer allow any advertising on electoral, political and social issues in the EU. This includes advocacy ads, meaning civil society groups risk losing a crucial channel. In comparison, well-resourced anti-rights forces and states that spread disinformation have many other ways of shaping public debate. Meta’s decision deepened this inequality of communication, demonstrating the power tech companies have over public discourse and their willingness to push back against regulations.
Meanwhile, Australia took a drastic step that others may follow, banning under-16s from using major social media platforms such as Instagram, TikTok and YouTube. The measure is controversial: proponents, including the government and child safety campaigners, argue it will protect children from harmful content, safeguard their mental health and prevent bullying; those against see danger in the government deciding who can access information, and raise concerns that ID checks create privacy risks and could enable data misuse. They also argue the ban could isolate vulnerable young people, particularly young LGBTQI+ people. It will be vital to assess what works and what doesn’t before others follow suit.
Civil society is working to defend itself from online threats. Digital security practice is improving, and civil society organisations have developed support tools. Activists routinely use VPNs to evade censorship and are adopting more secure platforms.
Philippine journalist and Nobel laureate Maria Ressa warns of ‘extremely dangerous times ahead’ following Meta’s decision to end its fact-checking programme on Facebook and Instagram. on 8 January 2025. Photo by Jam Sta Rosa/AFP
Civil society is combatting disinformation through fact-checking initiatives in many countries, including Argentina, Bolivia and Ecuador. In Malawi’s recent election, civil society used early warning systems to track online disinformation and hate speech, helping reduce tensions. In Nigeria, a civil society group has launched a WhatsApp chatbot allowing users to verify information.
Civil society is also working to hold tech companies to account, including through litigation. Meta and its outsourcing partners face three court cases in Kenya accusing them of promoting hate and incitement during Ethiopia’s conflict. An Ethiopian man whose father was killed after being doxxed is among those accusing Meta of failing to prevent amplification of hate speech. The cases seek improved content moderation policies and a compensation fund.
Many more such actions can be expected. In a fast-evolving sphere that affects every aspect of human life, civil society must stay engaged, continually refreshing its understanding of and responses to technology. Civil society must keep exposing the unaccountable and irresponsible power of tech oligarchs and upholding the human rights standards states and tech companies evidently think aren’t needed.