Receive free AI updates
we will send you myFT Daily Digest Email summary of latest information AI There is news every morning.
The debate over who should regulate AI has always been top-down. Tech giants say they want elected officials to set limits. But Washington has a hard enough time keeping up with targeted advertising and surveillance capitalism. U.S. states have AI regulatory proposals — often corresponding to large industrial use cases in their regions. European and Chinese authorities are also exploring ideas.
However, no one fully understands the capabilities of new technologies, which makes finding the perfect, purpose-designed solution difficult.
But one group has just made significant progress in building some new guardrails — the Writers Guild of America, which represents prominent Hollywood writers who have just struck deals to return to work. In addition to higher wages, residual funds, and minimum wage for employees, the authors also got something arguably more important: new rules about how the entertainment industry can and cannot use artificial intelligence.
These rules apply to any project using unionized writers, who can decide whether they want to use artificial intelligence in their writing. Studios must also disclose to writers whether any material provided to them was generated by artificial intelligence – which cannot be used to undermine the writers’ own intellectual property.
This is a big deal. First, it shows that AI can actually be regulated. While technologists like to act like they’re begging Washington to step in so their new products and services don’t blow up the world, the truth is they’re spending billions trying to draw a regulatory line that will let them get as much as possible Legal protections address possible problems as much as possible while also allowing them to continue to innovate. For CEOs, stakeholder concerns are far less important than keeping up with their peers in Silicon Valley and China.
The second reason this agreement is important is that these new rules are not being implemented from the top down, but from the bottom up. Workers who use new technology on a daily basis have a good understanding of how to contain it appropriately.
“Workers know something,” said Amanda Ballantyne, director of the AFL-CIO Technology Institute, who I discussed these developments with last week at the Code for Artificial Intelligence conference in Southern California. . “Unions have a long history of using the knowledge of working people to create better rules around safety, privacy, health and human rights.”
She noted that unions were critical to the rollout of other transformative technologies, such as electricity, helping to shape new industrial systems to improve safety and productivity. The success of the Tennessee Valley Authority project in the 1930s was largely due to the input of the International Brotherhood of Electrical Workers, an organization that grew in step with new technologies. The union has made a series of recommendations to the government on how best to organize a major project to electrify rural areas in the south. Unions were also key to the successful industrialization efforts in World War II and set some of the subsequent factory standards.
To the Germans and Japanese, the idea that workers “know something” is not surprising. In recent decades, both countries have adopted more collaborative workforce models to capture market share from the U.S. auto industry. Detroit is often criticized for not embracing Asian-style lean manufacturing methods sooner, but these systems rely on minute-by-minute collaboration between workers and managers, which requires trust — something often lacking in the United States.
Collective bargaining in the United States is contentious, and in some ways corporate America has the system it deserves—in the early days, companies chose to negotiate simply around pay, resisting production methods that involved sharing power. But that’s not necessarily the case with the relationship between workers and bosses when making decisions about new technologies like artificial intelligence. In fact, there is a strong argument that management should interview employees as new technologies are rolled out to understand which technologies help increase productivity, which ones undermine privacy, and which ones create new opportunities and challenges.
In the best case scenario, this could develop into a digital improvein which employees and management make incremental changes, slowly but surely working together to deepen their understanding of AI.
Most people understand that if AI is not anthropocentric and ultimately augments the human workforce, we are in for some very ugly politics. A recent academic study found that 80% of the U.S. workforce will have at least some job tasks changed by artificial intelligence. This is another reason to take a bottom-up approach to managing new technologies. Labor has day-to-day experience on the frontline of using AI and can help deliver the best skills training to ensure the new tool is a win-win.
Union-led regulation of artificial intelligence looks likely to spread. SAG-AFTRA, the union representing high-profile actors, is taking a closer look at writers’ AI deals, as are other labor groups. All this raises a broader discussion about unions as potential data stewards, protecting the interests of workers and citizens. In both areas, labor can serve as an effective check on big tech companies and big countries.
rana.foroohar@ft.com
