Orchid AI Backlash: 'Second Brain' Assistant Ad Sparks Fury

Silicon Valley has spent the past year promising that AI assistants will quietly take over the drudgery of daily life. In late July, a Y Combinator-backed startup called Orchid discovered exactly where that promise stops being welcome: at the border of a romantic relationship. A two-minute launch video showing Orchid's "second brain" chatbot rescuing a forgotten anniversary became one of the most-argued-about pieces of AI marketing of the year, pulling in millions of views and a torrent of comments calling it dystopian, disgusting, and worse710.

The ad that started it all

The setup is deliberately mundane. A young man is absorbed in video games while his anniversary slips past unnoticed. His partner, rather than confronting him, texts Orchid — and the assistant springs into action, booking a restaurant and ordering lilies, which it remembers she prefers to roses. The man, seemingly aware the machine has already handled everything, taps out a protest that he can manage this himself. Orchid's reply is the line the company clearly hoped would become its hook: "you literally can't. that's why I'm here," the whole exchange rendered in lowercase and emojis1011.

The company said the scenario was built "around a real situation," an explanation that arguably made things worse, since it implied that forgetting an anniversary under game-console duress is a software problem awaiting a software fix10. The video, posted on X on July 28, passed 3.2 million views within days, according to reporting that cited Fast Company's tally, and drew more than 600 replies — a striking ratio of fury to curiosity7.

The reaction on Instagram and X was unusually raw even by the standards of AI backlash. One commenter asked what the point of a relationship is if all the effort and compassion are outsourced; another called the product "nefarious"; a third declared that if it doesn't end humanity, nothing will10. Wired highlighted a viewer who said the ad seemed aimed at "adult babies uninterested in the world they live in," and one widely shared reply — "tech people need to stop" — captured the general mood7.

What Orchid actually sells

Here is where the story gets more complicated than the outrage suggests. Strip away the anniversary framing, and Orchid is a fairly conventional product for 2026: a text-message-based executive assistant that lives in iMessage, texts you an outline of your day each morning, drafts replies, and books meetings once you approve them7. It integrates with Gmail, Google Calendar, and Google Drive to manage inboxes, schedule calls, organize travel, handle invoices, and remember personal details. Y Combinator, which lists Orchid in its Spring 2025 batch, describes it as an executive assistant; the company's own materials pitch it at founders, investors, lawyers, and agency owners who want to offload administrative work, and its only published case study concerns helping a corporate executive prepare for client calls7. It was founded in 2025 by Nizar Abi Zaher and Adam Wazzan in San Francisco7.

Co-founder Nizar Abi Zaher has pushed back on the viral reading of the ad, telling Gizmodo that the anniversary scenario was meant to demonstrate a specific technical capability — letting multiple people communicate with the same AI agent and share conversation context — and that today that feature extends to group chats and coordinating meetings, not romantic salvage operations7. The company has also stated that it does not sell user data or use it to train models, and that drafted emails stay in drafts until a human sends them.

That gap between the marketing and the machinery is the real story. Orchid is not selling an anniversary robot; it is selling calendar and inbox automation, wrapped in an ad that deliberately wandered into emotional territory because emotional territory gets views. The company got the views. It also got a reputation problem it did not bargain for.

The allergen ad that made things worse

The anniversary spot was not even the most controversial thing Orchid produced. A second, since-deleted ad suggested the AI agent could scan your food and check it for allergens before you eat. The script's own language acknowledged the stakes — the list of things that put the speaker in the hospital is long enough that anything is the problem — which is precisely why viewers panned it, since trusting a consumer chatbot to vet food for someone with severe allergies is a safety claim of an entirely different magnitude than booking a table10[1112]... The since-deleted spot was quickly criticized for the obvious risks of positioning a convenience product as a medical authority4. On Reddit, allergy and gluten-free communities called the marketing "dangerous and deadly", and the watchdog account More Perfect Union amplified the critique on TikTok, noting the company had advertised the product as remembering a child's life-threatening allergy so parents no longer have to "remember what matters".

For anyone building consumer AI, this is the sharper lesson of the Orchid episode. There is a defensible line between an assistant that reduces cognitive load and one that presents itself as an authority on matters of physical safety, and Orchid appeared to sprint across it in pursuit of a memorable hook.

Privacy is the quiet half of the story

Beneath the cultural argument sits a practical one that has received less attention but may matter more. The viral ad's shared-agent premise — a bot texting separately with both partners while holding context between them — raises questions the company has not fully answered: what does the assistant report back to each person, how is data shared between connected accounts, and who controls which messages the AI can disclose7. Analyzing the preferences of one partner for the benefit of the other, without that partner's explicit consent, is a genuinely murky area under privacy regimes like the EU's GDPR, and a "second brain" by definition demands continuous access to intimate communications, financial data, and behavioral patterns. Orchid says it doesn't sell or train on user data, but the specifics of multi-user shared contexts remain ambiguous.

The adoption backdrop: consumers already treat AI as personal infrastructure

What makes Orchid more than a punchline is that it did not emerge from nowhere. Economists at OpenAI and Harvard analyzed 1.5 million ChatGPT conversations spanning 2022 to 2025 and found that roughly three-quarters of interactions involved non-work tasks — practical guidance, information seeking, writing, cooking advice, finding a pharmacy10. The dating app happn declared the "AI situationship" an emerging trend this year, with chatbots used as "emotional mirrors" and spaces for attentive listening1011. A study by the Austrian Safer Internet Centre found about a fifth of European adults have used AI chatbots for romantic conversations, and more than a quarter believe young people could fall in love with one11. Surveys cited in the coverage found 26 percent of U.S. singles used AI in their dating lives in 2025 — a 333 percent jump year over year — and 44 percent of married adults have turned to AI for relationship advice, climbing to nearly two-thirds among millennials. Professional assistants are skeptical too: on Reddit's executive-assistant forum, users dismissed Orchid as "a reminder app" whose marketing fluff targets people who could never afford a human EA.

Where the coverage converges — and my read

The reporting agrees on the facts and mostly agrees on the diagnosis: consumers will let AI automate workflows, but they react viscerally when it appears to automate care3. Where accounts diverge is on what Orchid is. Euronews frames it as Silicon Valley solving problems nobody outside the Valley has10. Archyde and the deeper startup coverage correctly note the actual product is an administrative tool for professionals, and that the ad oversold it7. AINave's analysis lands on the builder-facing takeaway: automate coordination before intimacy, and keep consequential actions reviewable.

My own reading is that Orchid's mistake was not the technology or even the ad's premise — it was a category error about what the gesture is for. A marriage therapist quoted by Wired warned that outsourcing difficult communication denies couples the "messy but meaningful growing pains" of a relationship, and flagged the triangulation risk of two people talking through a third party7. The deeper problem is authenticity: when both partners know the machine staged the flowers, the gesture is theater, and a man who forgets his anniversary but reaps the credit has been handed a machine for pretending to care3. Notably, the one part of the story consumers seemed to accept without complaint is the boring part — an assistant that manages inboxes, travel, and calendars. That split is the most useful signal the entire episode produced for the industry: consumer AI adoption is racing ahead on logistics and stalling on love, and the boundary between the two is where the trust — and the regulatory and safety questions — now live.