Google has seriously leaned into AI enshittification lately

Google used I/O to push deeper AI integration into Search, including expanded AI Overviews, AI Mode, and new ad formats embedded within AI responses. The article argues this could degrade search quality, obscure open-web links, and intensify user and developer backlash, especially alongside the closed-source replacement of Gemini CLI and concerns over local model installations on devices. The near-term market impact is likely limited, but the news reinforces growing antitrust, trust, and product-risk scrutiny around Google.

Analysis

Google is signaling a shift from traffic broker to answer-layer landlord, and that is structurally negative for the open-web ecosystem because it collapses the handoff between discovery and monetization. The first-order hit is to publishers and shopping affiliates, but the second-order hit is to the data sources that power search quality: once reputable sites see less referral value, they have less incentive to maintain crawl-friendly, structured content, which degrades model inputs over time and can create a self-reinforcing quality spiral. That dynamic is especially bad for GOOGL because the more the interface feels like a closed, commercialized recommendation engine, the more users will seek alternate search paths, ad blockers, and niche tools.

The bigger strategic risk is not near-term backlash; it is that Google's AI layer trains users to bypass the web and then forces advertisers to subsidize the compute. If those AI journeys become materially more expensive per query than classic search, margins compress unless ad load rises further — which increases user irritation and accelerates substitution. That is a classic enshittification trap: short-run ARPU support, medium-run engagement decay, long-run competitive opening for privacy-first or utility-first alternatives. The open-source tooling move compounds the issue by turning developers from contributors into critics, which raises churn risk in the ecosystem that has historically amplified Google’s product moat.

META and MSFT face a different but related problem: they are exposed to the same user rejection of AI clutter, but their businesses are more diversified and less tied to search intent purity. META is vulnerable if AI-generated content and ad injection worsen trust in feed quality; however, it also has the option to copy the format and spread the monetization burden across a larger inventory base. MSFT’s risk is more subtle: enterprise AI adoption may continue, but if model-driven workflows prove slower or costlier than promised, the market will eventually compress premium AI multiples across infrastructure and software beneficiaries. The contrarian view is that the market may be underestimating Google’s willingness to absorb user dissatisfaction if even a small lift in commercial intent offsets traffic loss; that supports near-term earnings but worsens the long-duration franchise value.

Catalysts matter on a months horizon, not days: watch for developer backlash, browser-side defaults, and any evidence that AI answer pages lower outbound clicks or increase search abandonment. The cleanest tell will be whether alternative search products or ad-blocking usage rise materially; if they do, the business is not just changing format, it is losing control of the demand funnel. If regulators start treating AI answer insertion as a material presentation/deception issue, the timeline to forced concessions could compress to 6-12 months.

Key Decisions for Investors

• Short GOOGL vs long QQQ for 3-6 months: thesis is not collapse in revenue, but multiple compression as investors price higher traffic-friction, rising compute cost, and worse brand trust. Cover if search engagement metrics hold while AI monetization proves accretive.
• Buy GOOGL put spreads 3-6 months out on any post-event bounce: structure for defined risk because the core issue is franchise erosion, not a single-quarter miss. Best risk/reward if implied vol stays below realized volatility.
• Pair trade long MSFT / short GOOGL: MSFT has AI exposure with enterprise distribution and less existential dependence on user trust in a single interface. Use as a relative-value hedge against a broader AI re-rating unwind.
• Reduce META exposure on strength if AI feed tooling expands ad clutter further: downside is not immediate revenue collapse, but higher user fatigue and lower time quality over 2-4 quarters. Favor a staggered exit rather than an all-at-once sale.
• Watch RDDT as a tactical beneficiary on any search trust wobble: if users migrate toward human-sourced discussion for product and problem-solving queries, RDDT can see incremental attention. Use only as a trading overlay, not a core long, because the platform is also exposed to AI-content contamination.