Damien Charlotin is a legal technologist and researcher associated with HEC Paris and Sciences Po Paris and creator of the global database of AI Hallucination Cases.
We discuss AI’s impact on legal work and the EU’s digital policies.
The bottom line: the use of AI in legal work is very beneficial to both lawyers and society, while the EU’s policies on AI and privacy are definitely not.
Here are my own takeaways from our conversation:
First, the use of AI in legal works is highly beneficial for lawyers and society, because it allows smaller law practices to take on big cases, generally improves access to legal knowledge, helps with data management in very complex cases, and automates trivial, non-interesting parts of lawyers’ work, such a renaming files. Overall, AI will increase access to legal knowledge at a much cheaper price, which means that qualified legal advice will be in high demand. Ultimately, Damien believes that AI will lead to more work for lawyers, not less.
Secondly, EU’s strict regulation of data and AI often leads to a meaningless compliance burden on companies, achieving nothing and making no one happy. There is a well-known emphasis on “data privacy” in EU’s regulation and if we ask people if they care about their privacy, most will say “Yes”. However, people’s actions suggest otherwise. Everyday, we gladly give away our behavioral data to American companies in exchange for free services. Because we genuinely get something good out of it. Almost all people are willing to sacrifice privacy in exchange for free, useful internet services, even if they say their privacy matters to them in the abstract.
The real threat is not abstract notions of online privacy, but a cyberattack from a malicious actor who steals a company’s customer data, trade secrets, sensitive business information etc. We rely on the tech giants to supply the structural security we take for granted. New, small tech companies in Europe (and other places in the world) don’t have the same levels of operational maturity, knowledge and experience to prevent attacks from malicious actors. Strengthening defensive cyber capabilities and ensuring a meaningful protection for users should be the regulatory focus, rather than forcing companies to go through comprehensive paper exercises about how they process data and protect people’s privacy.
‘Digital sovereignty’ is important in principle to ensure EU’s competitive relevance in future. Unfortunately, EU policymakers are unwilling to back up the high-flying buzzwords with action. For instance, the EU could offer an attractive prize to whichever contender that comes up with a viable alternative to Microsoft 365. But no such prize is on the table. Instead, the EU is going through a stage of moral panic about its dependency on tech giants and condemn the American products and services that people and businesses are deeply dependent on. The condemnation and moral outrage is satisfying, namely to Europe’s intellectual elite, but it doesn’t accomplish anything by its own or change reality.
The reality: People are not leaving Facebook, Instagram or TikTok in droves, most companies rely on OpenAI and/or Claude, and the market for Tesla cars is rebounding after the sales briefly plummeted when Musk was often in the news. There is a good reasons for this: the American services are frankly much better, than anything we can offer in Europe, people love them - even if they complain, and currently there is no meaningful indication that this could change in the near-term future.
Please join the conversation and share your thoughts in the comments.
Links:
Global database of AI Hallucination Cases
What jobs will AI destroy? Exhibit A shouldn’t be on the list. (Damien Charlotin/Washington Post)
