High-quality data is the bottleneck of AI innovation in both enterprises and academia, a neutrally-governed open data community could be the key to solving this Community-led efforts to solve 3 critical problems in open data: 1.Dataset format: an interoperable data structure. 2.Dataset standard: unify open dataset standards in collecting, sharing and exchanging to promote open data collaboration and reduce liability risks 3.Dataset licensing: standardize data lineage and license review process to reduce contributors and distributors’ liability risks