This month, RAW researcher Marijn Hoijtink contributed to two pieces in the Belgian press that speak directly to RAW’s research on how military AI is being developed, justified, and deployed.
In Knack, Hoijtink points to a significant shift in how military technology is being procured and imagined. Where investment in military tech once aimed at precision, striking more accurately and removing the “dirty work” from soldiers, companies now want to use it to operate at massive scale and overwhelm adversaries. Under Pete Hegseth, the US administration has embraced this logic fully, abandoning the classic model of extensive weapons testing in favour of speed. The risk, as Hoijtink notes, is that belief in technological dominance leads armies to overestimate their own capabilities and prioritise tactical gains over long-term strategy.
In De Morgen, she turns to the European dimension, noting that governments are partnering with companies like Palantir out of urgency, eager to scale up and embrace new technologies. But innovation in defence means embracing disruption and risk, which sits uneasily with the need for certainty about what weapons actually do. That urgency, she observes, is partly fed by narratives, like the idea that China will attack in 2027, that are in her words, “completely unfounded,” yet shape political and corporate discourse alike.
When speed becomes the organising principle of weapons development, questions of accountability, oversight, and civilian harm don’t disappear, they just go unanswered.
