‘Big Short' trader Michael Burry slams OpenAI, Anthropic for ‘self-serving' calls to slow AI – joining backlash

Michael Burry criticized OpenAI, Anthropic and major AI platforms for advocating slower AI development, arguing the push is self-serving regulatory capture designed to mask decelerating growth and support eventual IPO valuations. OpenAI and Anthropic reportedly filed IPO paperwork earlier this summer, but Sam Altman said a listing would be ill-advised amid current AI concerns. The dispute has added pressure to AI-linked equities as executives including Anthropic's Dario Amodei, Altman, Elon Musk and Google DeepMind's Demis Hassabis endorse more deliberate development and government oversight.

Analysis

The immediate equity risk is not a binding AI rule but a narrative shift from capability scarcity to monetization scrutiny. MSFT and GOOG can absorb compliance costs and may ultimately gain from rules that raise entry barriers, but their near-term multiples remain more sensitive to evidence that enterprise AI revenue is lagging the pace of infrastructure spending. Treat executives’ safety messaging as strategically motivated rather than as independently verified evidence of a material product-development slowdown.

Over the next 1-3 months, the key read-through is whether investors begin discounting lower frontier-model training intensity: that would pressure the AI-capex complex more than platform owners, while improving free-cash-flow expectations for MSFT and GOOG if reduced spend does not impair cloud demand. A voluntary federal framework is unlikely by itself to change earnings, but detailed testing, reporting, or licensing thresholds could favor incumbent labs and hyperscalers at the expense of smaller model developers and venture-backed challengers. The more consequential catalyst is upcoming cloud commentary on AI revenue conversion, backlog, depreciation and capex guidance.

The contrarian view is that a coordinated moderation narrative is bullish for the incumbents’ economics, not necessarily bearish for AI: less model-race spending can widen returns on capital while compliance requirements entrench distribution advantages in Azure, Google Cloud and enterprise software. This thesis fails if customers view slower model releases as evidence that productivity gains have plateaued, driving weaker cloud consumption and renewed capex cuts. Watch for AI-related cloud growth deceleration, rising depreciation as a percentage of revenue, or guidance that capital intensity will remain elevated without corresponding AI revenue disclosure.

Key Decisions for Investors

• Maintain a modest 3-6 month long MSFT / short equal-dollar AI infrastructure basket exposure rather than a directional AI short; the trade benefits if frontier spending moderates while Azure distribution monetizes. Use a reversal in MSFT AI/cloud growth or a material increase in capex guidance without revenue conversion as the stop condition.
• Prefer GOOG over smaller, capital-constrained private-model proxies for a 6-18 month regulatory-barrier thesis; Google can internalize model-evaluation and compliance costs while retaining Search, Cloud and distribution optionality. Do not add aggressively until the federal framework’s thresholds and enforcement mechanism are public.
• Avoid treating the safety-policy debate alone as a catalyst to short MSFT or GOOG over the next several days; the financial signal is weak absent evidence of lower enterprise AI demand or reduced cloud backlog. Set an earnings-season alert for explicit AI revenue, cloud consumption and depreciation disclosures.
• For portfolios with concentrated AI-beta gains, buy 3-6 month downside protection on broad technology exposure rather than single-name puts: the principal risk is a sector-wide multiple reset if management teams acknowledge slower monetization, while individual hyperscalers may be partially protected by regulatory capture.