As artificial intelligence (AI) becomes embedded in everyday software development, technology companies are still divided on whether candidates should be allowed to use AI tools during technical hiring rounds. While some firms see AI-assisted coding as an inevitable workplace reality, others maintain that it obscures a candidate’s core problem-solving ability, leading to sharply differing interview practices across the industry.
The lack of consensus is evident even among AI-first companies. Anthropic, the developer of Claude, had barred candidates from using AI tools or chatbots during interviews as recently as May, 2025. The company later in July that year revised its stance, allowing limited use for resume preparation but not during coding assessments unless explicitly stated. Last week, Tristan Hume, who leads performance optimisation at Anthropic, said the company had redesigned its take-home test yet again as newer models continued to outperform human applicants.
“Each new Claude model has forced us to redesign the test,” Hume wrote. “When given the same time limit, Claude Opus 4 outperformed most human applicants. That still allowed us to distinguish the strongest candidates – but then, Claude Opus 4.5 matched even those.” The challenge, he noted, is identifying what value a developer adds when AI generates much of the code.
Intuit’s focus on core thinking
At Intuit India, the use of AI tools is encouraged on the job but restricted during interviews. Nidhi Gopal, vice-president for mid-market and desktop, said teams routinely use tools such as Cursor, Qodo and Windsurf, helping them deliver code about 40% faster. “But during interviews, we prefer to evaluate the candidate’s core intelligence and how they apply their own thinking to design and coding fundamentals,” she said. The company focuses on problem decomposition, communication and the ability to improve AI-generated code, rather than raw output.
Some startups are taking a more situational approach. Bengaluru-based voice AI firm Gnani AI discourages AI tools for junior roles but allows limited use for senior positions such as tech leads or architects. “At junior levels, we want to assess raw problem-solving and analytical skills,” co-founder and Chief Product and Engineering Officer Bharath Shankar, said. “For senior roles, where system design is key, AI tools can support the discussion.” He added that candidates are expected to justify their choice of language, algorithms and approach, often through real-world prototype building.
Global tech giants remain split
Globally, policies remain split. Amazon has imposed a blanket ban on AI tools during interviews to ensure a level playing field. Meta, in contrast, began piloting AI-assisted coding rounds in late 2025, allowing candidates to use multiple models within a monitored environment. Canva has also reworked its hiring process, replacing traditional coding questions with more open-ended, AI-assisted challenges designed to test engineering judgement rather than syntax recall.
“The shift is clear,” Sashi Kumar, managing director at Indeed India, said. “As AI becomes part of everyday development, employers are increasingly evaluating how candidates reason through problems, validate AI output and make trade-offs around performance and security. Coding ability is no longer about speed alone, but about judgment and responsible use of AI,” he added.