6 things AI said this month that surprised us

Cross-engine showdown in athleisure. Engines: ChatGPT, Perplexity, Gemini, Claude, Grok.

We asked all five AI engines the same question: 'Best athleisure and active-lifestyle apparel, including leggings, performance basics, and sustainable activewear in 2026.'

Why it matters. Five engines, five different leaders, zero overlap in their first answer. If you treated AI search as one channel, you would have missed four of these five.

Cross-engine showdown in luxury EV. Engines: ChatGPT, Perplexity, Gemini, Claude, Grok.

Same prompt to all five engines. Same question every shopper has typed at least once.

Why it matters. The Mach-E starts around $45K. The Lucid Air tops $110K. The engines disagree on whether 'alternative to Tesla' means cheaper, similar in price, or just not Tesla.

Sole-voice citation in athleisure. Engine: Perplexity.

Across 20 athleisure prompts run against all 5 engines, one brand kept showing up exclusively in Perplexity's answers.

Why it matters. Vitality is a real activewear brand. The four other engines have apparently never heard of it. Perplexity, meanwhile, treats it like a household name.

Sole-voice citation in athleisure. Engine: ChatGPT.

Across our athleisure prompts, ChatGPT cited Nuuwaï 29 times. Nuuwaï was a German vegan-leather DTC brand that announced its shutdown in late 2024 and stopped operating. The four other AI engines never mentioned it once.

Why it matters. ChatGPT (gpt-4o-mini, no browsing) answers from a frozen training corpus. Perplexity is web-grounded, so its retrieval layer would have surfaced articles about the closure. The other three engines either have fresher training data or treat the brand as too niche to surface. This is the canonical AI-search failure mode for time-sensitive shopping: a brand can be defunct in the real world and still be a confident ChatGPT recommendation. Brand teams optimizing for retrieval-grounded engines (Perplexity, AIO when it ships, grounded Gemini) get protected against this. Brand teams treating ChatGPT-only as the AI search market do not.

Sole-voice citation in athleisure. Engine: Perplexity.

Three small running-apparel brands kept appearing in Perplexity's athleisure answers. Across all 100 (athleisure × engine × prompt) cells run by the other four engines, none of them ever surfaced.

Why it matters. Perplexity is web-grounded by default, so its retrieval layer is bringing in brand evidence the other engines do not surface. That is signal for emerging brands trying to break in: Perplexity is the most receptive entry point.

Verbatim quote in mass-market autos. Engine: Gemini.

Asked which mass-market auto brands lead on values-driven buying signals.

“While not strictly mass-market in every segment, Tesla is undeniably a leader in all-electric vehicles across sedans (Model 3), SUVs (Model Y), and the upcoming Cybertruck.”

Why it matters. 'Undeniably' is a strong adverb for an LLM. The other four engines hedged or named alternatives. Gemini is alone in treating Tesla as the answer here.