It’s worth repeating — I am not a lawyer. Personally, I believe that Judge Mehta and his team were technically competent, conscientious, and even courageous at times. Google also wisely pulled back on exclusivity clauses post-verdict, which gave the appearance of compliance.
I am old enough to remember both the Microsoft and AT&T (“Ma Bell”) antitrust suits, and I am sympathetic to the difficulties of predicting how any remedy might impact the industry or even the overall U.S. economy. It’s also, as the court noted, very difficult to determine how much of Google’s market dominance was fairly won with investment and innovation.
The court was particularly swayed by the massive disruption in search that GenAI/LLMs pose, potentially opening up new competition. Specifically, they noted that GenAI has attracted a surge of investment in search, stating that: “The money flowing into this space, and how quickly it has arrived, is astonishing.” I don’t think many of us would argue with that point.
Here’s where I disagree. I strongly believe that the court missed its own point regarding Google’s monopolistic advantages. While competitors like OpenAI have certainly disrupted the search market and are aggressively pursuing search-like capabilities, Google’s infrastructure, index, and user data remain massive advantages. Google still maintains an enviable pool of machine learning and AI talent, and has pioneered key innovations in the space, including groundbreaking work in transformers and self-attention that drove the GenAI revolution. GenAI competitors have massive hurdles to overcome in the search market, and I believe that Judge Mehta’s ruling may have suffered, like most of the industry, from AI hype.
Ultimately, we can’t expect the industry landscape to be rewritten by a single ruling, even at the end of a 5-year case of this magnitude. These remedies are serious, and I suspect Google will think twice about pushing too close to the line of another antitrust action. Even by internet-industry standards, the next 2–3 years in search and GenAI seem impossible to predict, but I believe this ruling will become a noticeable weight on the shifting balance.