Apple's Alignment Warning, and Google's AI Foothold in Schools
How this was made Verified AI
Every Intellegix briefing is generated from that day's broadcast and run through automated checks before it publishes — with a human paged on any flag. Here is the trail for this edition.
Apple's annual 10-K filing contains a disclosure that AI safety researchers will recognize immediately: the company warns investors that AI agents might not follow instructions. In plain language, systems Apple is building to act autonomously on users' behalf might take actions their creators did not intend — what researchers call goal misgeneralization or instrumental convergence. For agentic systems with access to a user's email, calendar, files, and applications, the potential for unintended consequences is non-trivial. By placing this warning in an SEC filing, Apple has simultaneously created a liability-protection paper trail and signaled to regulators that the risk is taken seriously enough to disclose formally — creating pressure on competitors who have not made comparable disclosures.
The Google schools controversy runs along a parallel track, concerning how AI tools are deployed in contexts where users are especially vulnerable and power dynamics are asymmetric. After Democrats in Congress accused Google of coercing schools by requiring them to accept Gemini alongside Search, Google responded that the two products have 'separate controls.' The catch: school Search still includes AI Overviews — Google's AI-generated summary feature — even when Gemini is turned off. The unbundling is partial at best.
The antitrust concern embedded in that arrangement is specific. Under Sherman Antitrust Act doctrine, having a large market share is not itself illegal — courts have been precise about distinguishing market dominance from monopolization. What is illegal is using a dominant position in one market to advantage products in an adjacent market through means other than competing on the merits. The allegation against Google is not that it dominates school-based Search, but that it is using that dominance to establish AI assistant adoption — the kind of conduct that triggers Section 2 scrutiny. Google is already in active antitrust litigation on its core search business, making the scrutiny here well-founded.
The educational equity dimension adds urgency. Schools in lower-income districts often rely almost entirely on Google's education suite. If that suite comes pre-loaded with AI features administrators haven't explicitly chosen, those communities become de facto test beds for AI tools without meaningful consent — and the habits formed among students shape consumer behavior for decades.