Shaping the Deep Tech Revolution in Agriculture 2025

Integration of CRISPR: Before and after FIGURE 12 Drivers and barriers to the use of CRISPR for agriculture FIGURE 13 Source: Consultations with AI4AI community experts AFTER The development of new cr op varieties r elies on traditional breeding and random mutagenesis. Developers cr oss-br eed plants over multiple generations to combine desirable traits such as dr ought tolerance or pest r esistance. This pr ocess takes a long time, perhaps several years, and can also introduce unwanted traits alongside beneficial ones. Even with marker -assisted selection of genes, pinpointing and editing specific genes is slow .With CRISPR, scientists can edit plant DNA with accuracy . AI facilitates this by pr edicting the best outcomes and simulating them befor e laboratory work begins. Instead of enduring long periods of trial and err or, scientists can accelerate optimal edits. This appr oach allows for the development of new cr ops in shorter periods. For farmers, this enables easier access to region-specific varieties that thrive under local conditions, boosting r esilience, food security and incomes.BEFORE DRI VERS Precision and speed: Moder n CRISPR-based tools allow for highly specific edits in much less time than conventional br eeding. This dramatically shortens development cycles for desirable traits. Declining R&D costs and improved accessibility: The cost of synthesizing guide RNAs (short sequences of RNAs that guide CRISPR enzymes to target spots for gene editing) and Cas enzymes (pr oteins that ar e used to cleave DNA) has fallen in the past decade, impr oving access to gene-editing platforms.Approval processes from laboratory to markets: The appr oval pr ocesses for gene-edited cr ops r equir e long periods of testing and ar e normally time- consuming. This pr ocess is also cost-intensive. Public perception: Public per ception and information asymmetry ar e critical barriers to the acceptance of genetically edited agricultural pr oduce. Complex traits and limited delivery methods: Several plant traits ar e contr olled by multiple genes and hence ar e complex to edit. BARRIER S Integration with high-throughput phenotyping and AI: Automated phenotyping platforms (tools used to analyse traits), coupled with AI-driven trait pr ediction, are allowing for rapid scr eening and validation of editing lines. This accelerates cr op-br eeding. Shaping the Deep-Tech Revolution in Agriculture 21