澳大利亚主要政党正逐渐走向罕见的共识。工党和联盟党越来越希望在人工智能领域押下重注。然而，就在政党们走到一起之际，澳大利亚民众却在背道而驰。
Australia’s major political parties are edging towards rare consensus. Labor and the Coalition increasingly want to take a major bet on artificial intelligence. Yet, just as the parties are coming together, the Australian community is running the other way.
去年，仅有30%的澳大利亚人认为人工智能的益处大于风险。到了2026年，公众对人工智能的信任似乎进一步下滑。
Last year, only 30% of Australians believed the benefits of AI outweighed the risks. In 2026, community trust in AI appears to be falling even further.
联邦议会成立人工智能联合特别委员会，这是不同寻常的一步。这表明，当前关于数据中心以及我们如何采用人工智能的决策，可能会在未来几代人中定义我们的经济、环境和社会。
It was an unusual step for our parliament to form a joint select committee on AI. It signalled that the decisions being made now on datacentres and how we adopt AI could define our economy, environment and society for generations.
随着新一轮听证会于本周开始，委员会成员必须面对人工智能政治与残酷现实之间的鸿沟。商界某些领域敦促我们将创新和人工智能采用置于优先位置，这往往是另一种说法：“更好地推销人工智能的好处。”无论你是人工智能乐观派还是怀疑派，这都是个糟糕的主意。萨姆·奥特曼、达里奥·阿莫代伊以及硅谷的其他巨头拥有金喇叭来兜售其技术的益处。澳大利亚人不希望民选代表也做同样的事——尤其是在该技术已同时带来机遇与危害之时。
As a fresh round of hearings starts this week, the committee members must confront the gulf between the politics of AI and the stark reality on the ground. Some parts of the business sector are urging us to prioritise innovation and AI adoption, which is often another way of saying: “Do a better job of selling the upside of AI.”Whether you’re an AI optimist or a sceptic, this is a terrible idea. Sam Altman, Dario Amodei and the other titans of Silicon Valley have a golden megaphone to spruik the benefits of their technology. Australians don’t want our elected representatives to do the same thing – especially when the technology is already bringing both opportunity and harm.
这并不意味着要对人工智能的益处视而不见。相反，澳大利亚是Anthropic的Claude及其他生成式人工智能应用人均使用率最高的国家之一。但我们也必须清醒地认识到潜在风险。至关重要的是，我们没有看到政府或人工智能公司本身对我们的担忧给出令人信服的回应。
This doesn’t mean being blind to the benefits of AI. On the contrary, Australia is among the highest users per capita of Anthropic’s Claude and other generative AI applications. But we must also be clear-eyed about potential risks. And, crucially, we don’t see a compelling response to our concerns from government or the AI companies themselves.
以人工智能对工作的影响为例。我们的社区正在提出切实的问题。人工智能将如何影响我们的工作，以及我们孩子的工作？政府正在采取什么措施，以确保澳大利亚工作场所使用的人工智能系统是安全的？
Take the impact of AI on work. Our community is asking practical questions. How will AI impact our jobs, and the jobs of our children? What is the government doing to ensure the AI systems used in Australian workplaces are safe?
我们的新报告《划清界限》阐述了人工智能如何推动有害的技术赋能型工人监控的兴起。工人们往往不知道自己被监控的程度——包括通过他们的工作设备和手机，以及通过微软 Teams 等软件进行监控。
Our new report, Drawing the Line, lays out how AI is propelling the rise of harmful tech-enabled worker surveillance. Often workers are unaware of the extent to which they are being monitored – including through their work devices and phones, and via software such as Microsoft Teams.
与联合人工智能特别委员会已经涉及的某些假设性（坦率说令人恐惧）风险不同，例如人工智能赋能的生物武器和工程化大流行，这种情况正在当下发生。
Unlike some of the hypothetical (and frankly terrifying) risks that the joint select committee on AI has already covered, such as AI-enabled biological weapons and engineered pandemics, this is happening today.
澳大利亚超市工人已经被迫满足人工智能生成的生产力指标，这些指标迫使他们在惩罚性后果的威胁下更努力、更快速地工作——使他们面临受伤和工伤事故的风险。
Australian supermarket workers are already being pushed to meet AI-generated productivity metrics that force them to work harder and faster under the threat of punitive repercussions – putting them at risk of injury and workplace accidents.
远程工作者已经遭受侵入式监控软件的侵害，这种软件能够监视他们在屏幕上做的一切——包括录制他们与家人交谈的音频。
Remote workers are already being subjected to intrusive surveillance software capable of watching everything they do on their screen – including audio recording conversations they have with their families.
网约车司机已经因与其表现无关的路况和客户投诉而受到惩罚，这可能导致他们的账户被停用并损失数天的工作。
Rideshare drivers are already being penalised for road conditions and customer complaints that are unrelated to their performance, which can result in their account being deactivated and days of work lost.
工人正在被悄无声息地常见的人工智能驱动监控工具压垮，这些工具已渗透进我们的工作场所和家庭办公室。而且影响是真实的。例如，一名年轻女性报告称，她遭到一名老板的性骚扰，该老板要求她随时开启视频。又如，仓库工人因机器的无情微观管理而产生自杀念头。
Workers are being quietly crushed by commonplace AI-powered surveillance tools that have crept into our workplaces and home offices. And the impact is real. Think, for example, of the young woman who reported being sexually harassed by a boss that required her to be on video at all times. Or warehouse workers driven to suicidal ideation due to relentless micromanagement by machine.
我们的职场法律无法保护工人免受不合理监控。它们也无法让雇主明确哪些类型的监控是合法的。
Our workplace laws don’t protect workers from unreasonable surveillance. Nor do they give employers clarity about what types of monitoring are lawful.
与人工智能带来的生存风险不同，后者不在澳大利亚政府的直接控制范围内，这个问题可以通过直接的法律改革来解决。
Unlike existential risks from AI, which are outside the Australian government’s direct control, this problem could be addressed through straightforward law reform.
我们敦促政府更新《公平工作法》，要求任何员工监控都必须“公平合理”。这建立在政府承诺要求《隐私法》涵盖的所有个人信息都得到公平合理处理的基础之上。
We urge the government to update the Fair Work Act to require any worker monitoring be “fair and reasonable”. This builds on the government’s commitment to require all personal information covered by the Privacy Act to be handled fairly and reasonably.
澳大利亚工人期待政府采取行动，保护他们免受人工智能危害。目前已有明确的改革方案可供采用。这并非假设的未来风险——让我们现在就解决它。
Australian workers expect the government to act to protect them from AI harms. There is a clear reform solution at hand. This is not a hypothetical future risk – let’s address it now.