Google Research 将联邦学习迁入 TEE,Gboard 支持外部可验证差分隐私
Google Research Moves Federated Learning Into TEEs: Gboard Now Trains With Externally Verifiable Differential Privacy
Google 把联邦学习的训练过程搬进可信执行环境,还允许外部核查差分隐私,Gboard 的英语和日语预测已经在用了。
Google Research 推出一套联邦学习系统,把梯度计算从手机转移到经过远程证明的服务器端 TEE 中执行。
访问策略发布到 Sigstore 的 Rekor 日志,二进制可复现构建,因此中心化差分隐私可由外部独立核查。该系统已在 Gboard 的英语和日语下一词预测任务上投入使用。
Google Research Moves Federated Learning Into TEEs: Gboard Now Trains With Externally Verifiable Differential Privacy
Google Research has unveiled a federated learning system that moves gradient computation from phones into attested server-side TEEs. Access policies are published to Sigstore's Rekor log and the binaries are reproducibly buildable, so central differential privacy can be checked externally. Gboard already uses it for English and Japanese next-word prediction. The post Google Research Moves Federated Learning Into TEEs: Gboard Now Trains With Externally Verifiable Differential Privacy appeared first on MarkTechPost .
本条情报归属 AITOP「产品」核心栏目,重点聚焦北美前沿实验室与硅谷科技巨头。在跨区域产业观察坐标中,该动态直接关联「AI 开发者工具链、Coding Agent 与工程化实践」研究主线,反映出当前技术演进与商业落地的最新风向。涉及核心实体与技术议题包括:Google、联邦学习、差分隐私、TEE,读者可通过下方标签进入对应专题页查阅历史脉络与关联对比。
信源核验说明:本文原始信息发布自一手权威信源「marktechpost」(Tier-1 一手发布),具备高确定性参考价值。