澳大利亚的AI透明度授权缺乏执行,使得大多数机构没有遵守。 Australia’s AI transparency mandate lacks enforcement, leaving most agencies noncompliant.

2025年的一项研究显示,只有大约45%的澳大利亚联邦机构有可查阅的AI透明度声明,尽管政府授权大多数机构在2025年2月之前公布这些声明。 A 2025 study reveals that only about 45% of Australian federal agencies have accessible AI transparency statements, despite a government mandate requiring most to publish them by February 2025. 许多声明难以找到或完全缺失,政策缺乏执法、惩罚或中央登记处。 Many statements are hard to find or missing entirely, and the policy lacks enforcement, penalties, or a central registry. 这种薄弱的监督削弱了公众的信任和负责任的人工智能的使用,特别是因为私营部门采用的道德做法仍然很少。 This weak oversight undermines public trust and responsible AI use, especially as private sector adoption of ethical practices remains low. 相反,美国和英国强制强制实行具有法律后果的AI登记制度,凸显了澳大利亚的 " 软触摸 " 方法存在的差距。 In contrast, the U.S. and UK enforce mandatory AI registers with legal consequences, highlighting the gap in Australia’s “soft-touch” approach.