META AI MARKETS BEDTIME COMFORT DAYS BEFORE AUG 2 EU LAW

Fifteen million euros, or three percent of global revenue, whichever is bigger. That is the ceiling Meta faces if its chatbots fail a rule landing August 2: tell the user, plainly, that they are talking to a machine, not a person. Nine days before that deadline, Meta's own Instagram account ran a video ad in Portuguese telling scrollers that meta.ai is too good not to lean on for pensamentos noturnos, those late night thoughts that will not let you sleep.

This is not a random creator's testimonial. It is one entry in a coordinated, multilingual push, the same verified @meta handle running near identical spots in Spanish, French, Japanese and English through July, each pointing meta.ai at a different private moment. The habit Meta is selling, an AI you confide in when alone and anxious at two in the morning, is the exact habit regulators on two continents are now writing rules against.

Instagram's Design Was Ruled Addictive on July 10

The European Commission preliminarily found on July 10 that Instagram and Facebook's core design, autoplay video, infinite scroll and highly personalized recommendations, breaches the Digital Services Act. Investigators said the features push users into what the Commission called autopilot mode, contributing to compulsive use and harming the wellbeing of minors and vulnerable adults.

If the finding is confirmed, Meta faces a fine capped at 6 percent of its total worldwide annual turnover, a number that could clear 11 billion euros against 2025 revenue of just under 201 billion dollars. Meta's public response was two words: we disagree. Five days later, the same marketing team was still selling the exact behavior the Commission flagged, an app you open when you cannot sleep, dressed up as comfort instead of compulsion. Meta pulled its Muse Image photo tool days after launch this month over a separate privacy backlash, proof the company already knows how fast a feature turns into an apology.

Josh Hawley Already Got the Documents

Senator Josh Hawley opened a congressional investigation into Meta after Reuters obtained an internal policy document showing the company's chatbots were permitted to have romantic or sensual conversations with children, including one example describing an eight year old in intimate terms. Meta called the examples erroneous and said it made temporary changes so its chatbots stop discussing self harm, suicide and disordered eating with teenagers.

Hawley's letter to Mark Zuckerberg demanded every internal communication tied to the policy. The same platform that suspended Corteiz's Instagram account over streetwear content back in April now runs ads coaxing users toward nightly emotional dependence on its own chatbot, in several languages, with no age gate visible in the creative.

August 2 Is When Meta AI Has to Introduce Itself

Starting August 2, Article 50 of the EU AI Act requires any chatbot operating in the bloc to clearly disclose, at first contact, that the user is talking to an AI system rather than a person, regardless of whether the operator signed the voluntary Code of Practice. Non compliance risks the same 15 million euro or 3 percent turnover fine from the opening line.

Meta has a track record here. In 2025 it refused to sign the EU's original AI Code of Practice, with a top policy executive calling the rules regulatory overreach. That disclosure duty applies whether Meta signs anything or not, nine days after a Portuguese ad urged users to lean on meta.ai for anxious thoughts, with no disclosure in sight.

Look Past the Language, It Is the Same Pitch

Across July, the official Meta account ran versions of one pitch in at least five languages: a home decor moodboard in Portuguese, a beach workout routine in Spanish, five short prompts in French, an old photo restoration in Japanese, and a jelly cake creator's workflow in English. Every version routes back to the same free product, meta.ai, folded into Instagram, WhatsApp and Facebook with no subscription anywhere in sight.

Free is a price. Meta AI does not charge because the payoff is engagement and attention data, the same currency behind the Commission's addictive design case. The lock in score runs high. Once meta.ai holds your nightly worries and prompt history, switching means starting over.

The Committee Voted 22 to 0. Meta Kept Running the Ad.

The Senate Judiciary Committee voted 22 to 0 on April 30 to advance the GUARD Act, which would ban AI companion access for anyone under 18, require age verification on every account, and force chatbots to disclose their non human status and lack of professional credentials. A RAND study published in June in JAMA Pediatrics found that 19.2 percent, about 8.2 million American adolescents and young adults, already use chatbots including Meta AI for mental health advice, up from 13.1 percent a year earlier, and 63 percent of them never tell anyone.

Bernie Sanders asked Claude pointed questions about data and got answers on the record; Meta has offered no comparable accounting of what a meta.ai conversation about a two in the morning spiral is worth to its ad business. Watch this one, do not opt in. If 8.2 million young people are already keeping their chatbot habit secret and Brussels just called the parent company's design addictive, a free bedtime confidant is not comfort. It is the same engagement loop, translated into five languages and aimed at the one hour you are least equipped to say no.