META's stock surged last week, but investors shouldn't ignore the risks. Meta shares climbed last week as Wall Street became increasingly optimistic about the company's AI strategy. The stock was up about15% for the week and erased the year-to-date losses. Investors are betting that Meta's enormous spending on AI infrastructure, custom chips, top engineering talent, and next-generation models will lead to faster revenue growth, stronger advertising tools, and new revenue streams over the next several years. The market clearly believes Meta has positioned itself as one of the leaders in the AI race. But while investors were celebrating, Europe reminded everyone that even great companies face meaningful risks. The European Commission announced preliminary findings that Facebook and Instagram may violate the Digital Services Act because of what regulators call "addictive design" features, including infinite scrolling, autoplay videos, and recommendation algorithms that encourage users to stay engaged for longer periods. If the findings become final and Meta does not make sufficient changes, the company could face fines of up to 6% of its global annual revenue, along with potential changes to how its platforms operate across Europe. Meta has disputed the findings and says it has already implemented significant protections for younger users. This could amount to a fine of around $12 B, but the bigger problem I see is a potential hit to ad revenue if they must change their business practices. Europe is an important part of their business considering it accounts for about 23% of overall company sales. We also can't forget the legal liability Meta is facing in the United States, which could ultimately total as much as $1.4 trillion. That number may sound shocking, but it stems from multiple lawsuits brought by numerous states and plaintiffs. The first major cases are scheduled to go to trial in August, with California, Colorado, New Jersey, and Kentucky leading the way. The lawsuits allege deceptive business practices, and potential penalties range from $2,000 to $20,000 per violation. Given Meta's massive user base, those fines could accumulate rapidly if the courts rule against the company. Beyond civil penalties, the states are also seeking disgorgement of profits, which would require Meta to surrender profits earned from the alleged misconduct during the relevant period. If Meta performs poorly in these initial cases, another 25 states have similar lawsuits waiting in the wings, significantly increasing the company's legal exposure. There are already signs that these legal challenges carry real financial risk. New Mexico recently won a $375 million judgment against Meta, and a separate federal trial is scheduled to begin early next year. The AI opportunity is also far from guaranteed. Today, investors are rewarding companies that appear to be winning the AI race, but the competitive landscape is becoming more crowded every quarter. OpenAI, Anthropic, Google, Microsoft, xAI, and others are investing billions of dollars to develop better models and attract developers. Meta has responded aggressively by spending heavily on infrastructure and recruiting top AI researchers, but there is no guarantee those investments will generate returns that justify the enormous capital being deployed. A big problem is today's leader in AI can quickly become tomorrow's follower if innovation slows. I also believe that all of these companies will not succeed in this space, which will mean enormous amounts of wasted capital for the losers. Wall Street seemed to be focused almost entirely on Meta's AI upside last week, and that optimism may continue to drive the stock higher. But investors should remember that valuation is increasingly dependent on AI execution while regulatory scrutiny remains elevated. If AI spending fails to produce the expected returns or regulators force changes that weaken engagement, today's bullish narrative could change quickly. Meta remains one of the strongest companies in technology, but even great businesses are not risk-free. As investors, it's important to weigh both the opportunities and the risks, not just the headlines driving the stock higher today. The spring home sales season disappointed in June The spring home-selling season ended on a disappointing note. Through May, existing home sales had been showing signs of improvement, and many real estate professionals were becoming more optimistic about the housing market. However, June's data told a different story. The conflict involving Iran contributed to higher inflation expectations and pushed mortgage rates higher, weighing on buyer demand. Existing home sales fell 2.4% in June to a seasonally adjusted annual rate of 4.09 million homes, well below economists' expectations for a 0.7% increase. Despite the monthly decline, the longer-term trend remains somewhat more encouraging. Existing home sales were still up 2.8% compared with a year ago, suggesting that underlying demand has not disappeared. There continues to be pent-up demand from prospective buyers, but many seem unwilling to make such a large financial commitment while borrowing costs remain elevated, even as housing inventory continues to improve According to Freddie Mac, the average 30-year fixed mortgage rate was 6.43% last week. If mortgage rates remain near these levels, many prospective homebuyers may continue to delay their purchases, preventing a stronger recovery in the housing market. Another Hidden Cost of AI: Steel Most people know that the AI buildout has driven up demand for advanced computer chips, contributing to higher prices for smartphones, laptops, and other electronics. They also know that AI data centers require enormous amounts of electricity, putting upward pressure on utility rates as more power is diverted to support AI infrastructure. But there's another cost that receives far less attention: steel. Steel is a critical component of every data center. Industry estimates suggest that new data centers will consume roughly 1 million tons of steel annually, representing approximately $1.4 billion in demand. Steel is used throughout these facilities from the structural columns, roof joists, and roof decking to the server racks that house thousands of AI processors. This growing demand has ripple effects throughout the economy. Higher steel demand can contribute to increased costs for automobiles, household appliances, commercial buildings, bridges, and countless other products that rely on steel. The impact doesn't stop there. Steel production is one of the most energy-intensive manufacturing processes. A single electric furnace steel mill can consume anywhere from around 50 to 200 megawatts of electricity per day, competing for the same power resources as AI data centers. As both industries demand more electricity, utilities face increasing pressure to expand generating capacity. Ultimately, who pays for that increased demand? The answer is often the consumer. Higher electricity demand can translate into higher utility bills for households and businesses as utilities invest in additional generation and transmission infrastructure. In regions where electricity supply is already tight, the competition for power is becoming even more apparent. For example, PJM Interconnection, the nation's largest regional transmission organization, plans to begin conducting supplemental power auctions with electricity generators in September to help secure additional supply. Auctions reward the highest bidders, meaning electricity increasingly flows to those willing to pay the most. As large industrial users and AI data centers bid aggressively for power, consumers could face higher electricity prices if supply fails to keep pace with demand. AI will likely bring enormous productivity gains and economic benefits over the long run. However, it is also creating secondary inflationary pressures that extend well beyond semiconductors. Steel, electricity, construction materials, and other critical inputs are all experiencing increased demand, and those costs eventually work their way through the economy. As the AI revolution accelerates, these indirect costs are likely to become an increasingly important part of the inflation story. Inflation Is Cooling... But Don't Pop the Champagne Yet The latest CPI report was another encouraging sign that inflation is moving in the right direction. Headline CPI declined 0.4% in June, marking the largest monthly drop since 2020, while the annual inflation rate slowed to 3.5% from 4.2% in May. Core inflation, which excludes food and energy, was flat on the month and eased to 2.6% year over year. Much of the improvement was driven by a sharp decline in gasoline and broader energy prices. While this is welcome news, I'd caution against declaring victory over inflation. One of the biggest challenges with inflation is that it doesn't always show up in the headline numbers immediately. It often works its way through the economy in waves, especially when it comes to energy. A good example is my own pool service. My pool guy recently raised his prices, likely for two reasons: higher chemical costs and the increased cost of driving from house to house. Those are both directly tied to energy markets. Even if gasoline prices temporarily fall and help bring down CPI for a month, businesses often adjust prices more slowly because they have to account for prior cost increases and the uncertainty of where energy prices are headed next. That's why I think investors should remain cautious. The recent improvement in inflation was helped significantly by lower oil and gasoline prices following a temporary easing in geopolitical tensions. But with conflict in the Middle East once again threatening energy supplies and oil prices recently moving higher, that relief could prove short-lived. The trend is encouraging, and the Federal Re