Researchers report using artificial intelligence to design CRISPR-like “molecular scissors” that are functional in cells. According to Nature, the work involves creating synthetic nuclease enzymes with characteristics not previously observed in nature. Multiple outlets describe the approach as an expansion of the CRISPR toolbox, emphasizing that the AI-generated enzymes show activity inside cells and can perform gene-editing tasks. The reporting highlights Jennifer Doudna, a Nobel laureate associated with CRISPR technology, as participating in or engaging with the AI protein-design effort covered by the outlets. The studies referenced focus on engineering enzymes through AI-based protein design rather than modifying existing CRISPR systems directly. While the articles describe improved or boosted editing efficiency in the context of the newly designed nucleases, they do not present a single unified performance metric in the provided excerpts. Overall, the coverage converges on the same core finding: AI-designed CRISPR-like nuclease candidates that do not naturally occur are able to cut genetic material in cellular experiments, pointing to a potential route for developing additional or more effective gene-editing tools.