I disagree. Many humans are phishing in a different language than their native tongue, and LLMs are way better at sounding legit/professional than many of them. The best spear-phishing will still be humans, but AI definitely raises the bar.