Britain's Advanced Research and Invention Agency is channelling £54m into what amounts to an evolutionary accelerator for wild species. Fourteen teams will split an initial £30m over two years, deploying CRISPR gene-editing, RNA interference, and engineered skin microbiomes on species from ash trees to swallowtail butterflies. The pitch: conventional conservation — reserves, corridors, reintroductions — cannot keep pace with the speed of climate-driven habitat change, so biology itself needs a faster clock. The most striking projects illustrate the ambition. Exeter researchers will use CRISPR-Cas9 to boost insecticide resistance in swallowtail butterflies, a red-listed species being killed by the very chemicals sprayed to protect crops. Rothamsted Research plans to compress oak reproduction from decades to months, letting long-lived trees evolve drought and pathogen responses within a human career rather than across centuries. Boston-based Cultivarium, partnered with Imperial College and Queen Mary, will engineer bacteria living on amphibian skin to produce proteins that block chytrid fungus — the pathogen behind population declines in more than 500 amphibian species and 90 confirmed extinctions. The programme sits inside a live global argument. At the IUCN's world conservation congress last October, more than 90 NGOs pushed for an outright ban on synthetic biology in conservation, warning of irreversible effects on natural systems. The ban failed narrowly after 250 scientists signed an open letter defending the tools. The IUCN instead adopted its first global policy on synthetic biology, with stringent criteria. Aria says this policy is one foundation of its own ethical framework, and none of the 14 projects will release engineered organisms before the programme ends in 2030. Governance looks robust on paper. Marie-Claire Cordonier Segger, a Cambridge international law professor chairing the programme's ethics committee, says the decision to restrict work to secure labs was taken early. Any future field deployment would require approval from the Department for Environment, Food and Rural Affairs, which has tightly controlled genetically modified organism releases. But the history of publicly funded biotech suggests that once tools exist and show promise in the lab, the pressure to deploy escalates sharply — and the regulatory apparatus may not have kept pace. The honest scientific assessment is sobering. Chris Thomas, founding director of the Leverhulme Centre for Anthropocene Biodiversity at York, calls the chytrid fungus project a "particularly worthy target" but cautions that impact will likely be "modest" and difficult to measure. Some amphibian species are already developing natural resistance; making certain species within a pond more resilient inevitably affects untargeted species. His bluntest line: thinking these interventions will play an important role in preventing biodiversity loss "is probably pretty naive." Aria itself has an institutional credibility issue that shadows the science. The agency — Dominic Cummings' brainchild, created to fund "crazy" ideas — has drawn criticism for geoengineering grants and for directing UK taxpayer money to US tech firms and venture capital. Chair Matt Clifford was forced out this month over a conflict-of-interest appointment at Anthropic. None of this invalidates the biology, but it means every pound spent carries a heavier burden of proof than a less politically charged funder would face. The core tension is structural. These projects are genuinely generative — new capability that did not exist before — but they create a new category of systemic risk if governance fails at the deployment stage. The £54m is modest relative to the ambition. The real test comes in 2028-2030, when successful lab results meet the political pressure to move into the field, and the question shifts from "can we?" to "should we, and who decides?"
