// trust

Source: Ars Technica

Malicious actors are exploiting Google's ad network to distribute freeze-screen scams, targeting both Windows and Mac users through legitimate search results—a vulnerability that bypasses the company's advertiser vetting. Rather than tricking people into clicking obviously fake links, attackers are now buying their way into the trust infrastructure itself, making the attack vector indistinguishable from normal browsing for most users. The incident exposes a core tension in Google's business model: the company's incentive to maximize ad volume and minimize friction for advertisers directly conflicts with its ability to prevent sophisticated fraud that uses its own platform as distribution.

Source: Search Engine Journal

As AI abstracts away links, citations, and publication signals from search results, brands lose the feedback loops they've relied on to understand what resonates with audiences—and consumers lose the ability to verify where information originates. Marketers can't see which content types or topics drive engagement, while readers can't evaluate credibility. The content-to-consumer chain becomes less legible. Companies optimizing for AI visibility are playing to invisible judges, forcing a rethink of how content gets discovered and trusted when links no longer signal authority.

Source: The Verge

Meta is attempting a rebrand away from child safety failures, antitrust battles, and leadership questions by emphasizing the metaverse and AI, but consumer skepticism runs deeper than new product categories address. The company faces a structural problem: younger audiences have formative memories of Facebook as their parents' platform, while older Gen X and millennial users carry active resentment about privacy violations and algorithmic harms. Demographic-wide forgetting is unlikely. Meta's real leverage isn't messaging—it's network effects. Users stay because their social graph is locked in, not because they've absolved the company of past conduct.

Source: NYT > Business (paywall)

As voters increasingly use AI chatbots as a primary source for election information, these systems produce inconsistent and sometimes contradictory responses based on user framing—a problem that worsens existing information fragmentation. Unlike search engines that point to source material, AI systems package political information as authoritative natural language, making it harder for users to recognize unreliable or personalized outputs. This threatens the shared informational foundation elections depend on, particularly since the companies building these tools have minimal incentive to prioritize consistency over engagement.