虎嗅

EU Incorporates AI Literacy into Law: Companies Will Be Held Liable if Their Employees Cannot Use AI

原文:欧盟把AI素养写进法律:员工不会用,企业要担责

Summary in Plain Language

This AI workplace report from the European Union breaks away from the tired stereotype that AI will lead to widespread job losses. We’re not yet at a stage where mass layoffs are happening; instead, a quarter of all jobs worldwide are undergoing a “silent transformation” – the tasks you do every day are being rewritten by AI. The EU itself still owes from the previous wave of digitalization: it aimed to ensure that 80% of the population has basic digital skills by 2030, but we’re only at 60% so far, meaning we’ll likely miss that goal until 2036. Now, the introduction of AI has become a major challenge. As a result, the EU has reclassified AI literacy training from a company-provided benefit to a legal requirement. It also warns us that we can’t just focus on teaching employees how to use AI tools; we need to change training policies, company procedures, and clear paths for promotion and job transfers. Otherwise, the gap between those who adapt and those who don’t will only widen.

Detailed Explanation

1. Don’t Panic About Losing Your Job: It’s Your Tasks That Are Changing, Not the Jobs Themselves

Many people panic at the thought of AI affecting their jobs, imagining they’ll be laid off tomorrow. This report clarifies that in the 25% of jobs affected by AI, over 90% are being transformed, not eliminated. For example, tasks like writing drafts, researching information, organizing data, and sorting reports, which were once repetitive, will be taken over by AI. The freed up time should be used for more valuable tasks, such as understanding what leaders or clients need, verifying the reliability of AI results, and assessing the suitability of these results for the company’s context, as well as determining who is responsible for any issues that arise.

In China, many companies waste money on generic training programs for all employees. According to this report, we should break down tasks into those that will be automated and those that will become more valuable, and use the same budget more effectively. The EU’s situation serves as a reminder to all countries: if we don’t address the shortcomings of the previous digitalization efforts, the new wave of AI transformation will leave us behind.

2. AI Lowers the Bar for Doing Work, but the Most Valuable Skill Is Error Detection

People often think that the core skills in an AI-driven world are simply writing prompts and using tools, but this is a misunderstanding. The report uses the programming industry as a clear example: thirty years ago, programmers had to memorize complex instructions; now, with simpler languages, you can ask for an attendance system, and AI can generate the code for you. However, can you trust the code without testing it for errors or security vulnerabilities? The value of a programmer now lies in assessing the quality and maintainability of the code.

This logic applies to all industries. Financial workers need to understand the risks behind AI-generated data, and HR professionals need to spot biases in recruitment materials. Most AI courses focus on tool operation, but what’s truly scarce are skills in setting standards, making judgments, and ensuring quality – these are the skills that won’t be replaced by AI.

3. The Simpler the Tools, the More Important It Is to Use Judgment

The EU has made using AI a legal requirement, emphasizing the need for judgment. People often assume that using AI simply means knowing how to use tools, but this is misleading. For example, if AI generates a company announcement with incorrect information, the company could face significant penalties. Therefore, companies must train employees on how to use AI responsibly. Otherwise, the cost of mistakes could be much higher.

4. Training Employees in AI Alone Is Useless Without Changing Company Processes

Many companies waste money on AI training without altering their processes. If employees learn AI but the company’s old ways remain unchanged, the training is useless. For instance, if AI can generate 3 public account drafts in 10 minutes, and the KPIs are increased, it’s just adding more work. Skills must be integrated into company processes; otherwise, AI becomes a tool for exploiting employees.

5. Don’t Jump to a New Industry – Move Within Your Current Role

Rather than switching to AI-related fields from scratch, it’s better to make small changes within your current role. The report suggests “adjacent migration” – for example, an administrative staff member can become an AI system administrator. This approach avoids the uncertainty of starting from scratch and builds on existing skills, making it less likely to be replaced by AI.

In China, subsidies for digitalization often only cover software purchases, without supporting process changes and employee training. Without these complementary measures, the gap in AI capabilities between large and small companies will widen.

6. Focus on Building the Right Skills for the Future

The best way to prepare for the AI era is to develop skills that won’t be replaced by AI, such as judgment and understanding of boundaries. These skills are essential for maintaining your job in any industry. The EU’s approach highlights the importance of clear rules and processes to ensure responsible use of AI.