AI fears are being weaponised

What the report says

Al Jazeera opinion columnist Adji Bousso Dieng argues that fear around artificial intelligence is being used to shape policy in ways that protect the market power of major technology companies. Published on 17 October 2026, the piece says alarm over so-called existential AI risk is drawing attention away from nearer-term harms already linked to AI, including job displacement, higher energy use from data centres, disinformation, bias in algorithms and military targeting.

The article says the debate is increasingly framed around restricting open-source AI models, which are cheaper or free and can be adapted by users. In the author’s view, this could leave advanced AI development concentrated in a small number of companies and countries, a pattern she compares with forms of digital colonisation. The piece argues that the Global South is especially vulnerable because frontier systems are expensive to build and often do not fit local conditions.

As an example, the article points to Kenya’s AI-driven Social Health Authority rollout, saying imported assumptions in the system led to inaccurate income estimates and higher healthcare charges for poorer users. More broadly, it contends that developing countries need AI tools tailored to local needs in areas such as health, agriculture, energy and climate work, rather than only chatbots built for wealthy markets.

The author concludes that safety regulation should focus on testing, accountability and transparency rather than bans on open-source innovation. The article presents this as a debate over whether AI will remain decentralized and broadly useful, or become further concentrated in the hands of a few dominant firms.