CapcomA Phoenix Wright: Ace Attorney dev has responded after the iconic courtroom game was used to test AI models’ reasoning capabilities.
As reported by Automaton, Hao AI Lab used Capcom’s iconic detective game to test four AI models to discover their logic and reasoning capabilities. According to K.Ishi, the test used “Ace Attorney to evaluate the AI’s practical ability to find inconsistencies in testimony, select appropriate evidence to support them, and refute them most effectively.”
While none of the AI models beat the game, K.Ishi revealed that OpenAI and Google Gemini 2.5 Pro made it to Chapter 4. Meanwhile, Anthropic’s Claude 3.7 made it to Chapter 2, and Meta’s Llama-4 Maverick failed to complete Chapter 1.
However, following the findings, Ace Attorney developer and voice actor Masakazu Sugimori has responded to the research.
Ace Attorney dev responds to AI research
“I was shocked to find that the game he worked on had been used to test AI,” Sugimori wrote on X. “How should I put this, I never thought the game I worked on so desperately 25 years ago would come to be used in this way, and overseas at that. That said, I find it interesting how the AI models get stumped in the first episode.”
As for what he thinks of AI, Sugimori remains hopeful that both man and machine and coexist. “The possibility of games being involved in this is interesting…And at the same time, I think we humans will not lose to AI. Excluding individual ‘tasks’, I don’t think we can win in terms of work, so the role of humans is to think.”
It’s this reasoning to “think” that Sugimori believes is why the AI models used in Hao AI Lab’s testing struggled to beat Phoenix Wright. So, if you find yourself stumped on a particularly tough Ace Attorney court case, you’re best off using your own logic, that or using the old save scumming technique! (Couldn’t be me!)
