More companies are embracing AI in their hiring processes, encouraging job candidates to use artificial intelligence tools during interviews. Once wary of candidates secretly leveraging AI to craft strong responses, firms like Canva, Meta, and McKinsey are now shifting focus to evaluating how well applicants can work with AI—recognizing that the technology is now integral to real-world work. At Canva, the shift began when CTO Brendan Humphreys noticed that engineers were becoming significantly more productive using AI. Instead of banning the tools, the company flipped the script and invited candidates to use AI during technical interviews. “We just flipped the script and went, ‘OK, we're going to invite you to use AI,’” Humphreys said. The goal is to hire people who can not only write code but also thoughtfully guide AI to produce high-quality results. To ensure candidates aren’t just relying on AI for quick answers, Canva redesigned its interview questions to be complex, ambiguous, and open-ended—challenges that require critical thinking. Simply feeding a prompt into an AI tool won’t yield a strong response. Instead, the company asks candidates to share their screens so interviewers can observe how they interact with AI, assess its output, and refine results. “We want to see the interactions with the AI as much as the output,” Humphreys said. Similarly, Arcade, an IT infrastructure startup, now expects technical candidates to use AI tools during take-home assignments. CEO Alex Salazar said that since candidates were already using AI regardless of policy, the company decided to make it part of the evaluation. Candidates must submit transcripts of their conversations with AI, which helps assess not just their technical ability but also their judgment—what Salazar calls “taste.” He noted that the best outcomes come from iterative refinement with AI, not just one-time prompts. Meta is also adapting. According to prior reports, the company is developing a coding interview where candidates can use an AI assistant, a move meant to reflect real-world development environments. Meta argued that this approach makes AI-based cheating less effective and better prepares new hires for how they’ll actually work. McKinsey & Company is piloting a similar change, asking graduate candidates to use its internal AI assistant, Lilli, during case interviews. The aim is to gauge how well applicants can collaborate with AI tools in a professional setting. While companies welcome AI use, they’re not looking for candidates who misrepresent their skills. Executive coach Susan Peppercorn warned that even if someone gets past initial screening by over-relying on AI, they’ll eventually be exposed when asked to explain their reasoning. At Canva, Humphreys emphasized that the real test is understanding the thought process behind decisions—especially when using AI to generate code. “Now we’re testing for an ability to harness that power, to control that power—to kind of ride the dragon,” Humphreys said. The message is clear: in today’s tech landscape, the ability to work with AI isn’t optional—it’s essential.