Shares of major Indian information technology companies tumbled this week, with declines of as much as 7%, after US consulting and technology services giant Accenture trimmed its revenue outlook, reviving concerns about a broader slowdown in global IT spending. The selloff, reported by CNBC, hit a sector that has long been viewed as a bellwether for enterprise technology demand worldwide.

Accenture’s Warning Ripples Through the Sector

Accenture’s results and guidance are closely watched by investors in Indian IT services firms because of the deep linkages between the two markets — Indian firms count many of the same global enterprise clients as Accenture and often compete for similar outsourcing and digital transformation contracts. A cut to Accenture’s revenue outlook is typically read as a signal that corporate clients are pulling back on technology spending more broadly, and Indian markets reacted accordingly.

Renewed Growth Concerns

CNBC noted that the slump has fueled fresh concerns over sector growth, adding to a list of headwinds facing Indian technology exporters, including currency fluctuations, competition from AI-driven automation that could reduce demand for traditional outsourcing work, and softer discretionary IT budgets among Western corporate clients still adjusting to higher interest rates and geopolitical uncertainty.

Part of a Broader Global IT Spending Story

The Indian IT slump comes against the backdrop of an AI investment boom that is reshaping how enterprises allocate technology budgets. While spending on AI infrastructure and chips has surged — evident in the rally in semiconductor stocks that helped lift the Nasdaq nearly 2% this week, according to CNBC — that boom has not necessarily translated into stronger demand for the traditional IT services and outsourcing work that has historically been the bread and butter of large Indian technology firms.

Investors will be watching upcoming earnings from other major global IT services and consulting firms for confirmation of whether Accenture’s cautious guidance reflects a broader, sector-wide pullback or a company-specific issue.

Inside the sprawling halls of Paris’s Porte de Versailles, the atmosphere at the tenth anniversary of Europe’s premier technology gathering feels remarkably sober. The flashing holograms and robotic dogs of previous years have been quietly pushed to the periphery. Instead, the defining VivaTech AI trends centre on something far less cinematic: immediate, measurable commercial utility. Ten years since its inception, the conference has outgrown its adolescent fascination with what technology could do. Now, European founders and international investors are betting everything on artificial intelligence that actually works on the factory floor, in the back office, and across the supply chain.

This shift at VivaTech mirrors a broader correction across the global technology sector. The initial speculative frenzy surrounding generative models has collided with the harsh realities of corporate budgets and data privacy constraints. We have officially entered the deployment phase. Executives no longer want to pay for experimental software that hallucinates legal precedents or hallucinates customer service responses. They demand secure, ring-fenced tools that drive margin expansion.

The numbers reflect this systemic maturation. According to recent data synthesized by the Organisation for Economic Co-operation and Development (OECD), enterprise adoption of applied AI models is projected to drive a 1.4% annual increase in labour productivity across the Eurozone by 2027. Yet, the same dataset reveals a glaring friction point: only 18% of mid-sized firms have successfully integrated these models beyond pilot programs.

The gap between pilot and production is where the money is now being made. European venture capital has adjusted its focus accordingly. According to the Financial Times, funding for pure-play foundation model startups dropped by 22% in the first quarter of 2026, while capital allocated to vertical-specific AI applications surged. Investors are no longer funding the picks and shovels; they are funding the extraction.

Walking the convention floor this May, the changing guard is impossible to ignore. Startup booths are stripping the phrase Large Language Models (LLMs) from their primary marketing copy. The pitches have transformed. Founders are no longer selling the intelligence of their neural networks; they are selling automated invoice reconciliation, predictive supply chain routing, and immediate cost reduction.

Arthur Mensch, CEO of Paris-based Mistral AI, summarised this shift during a closed-door briefing on Tuesday. He noted that enterprise clients have abandoned open-ended experimentation in favour of strict, highly defined use cases. This pragmatism is fundamentally reshaping the European tech ecosystem. The continent, long criticised for failing to produce consumer internet giants, is leaning heavily into its traditional strengths: industrial engineering, regulatory compliance, and complex B2B software.

The capital backing these ventures is equally pragmatic. The French state investment bank, Bpifrance, announced a €500 million facility specifically earmarked for enterprise AI adoption within legacy manufacturing firms. This is not speculative capital. It is modernisation infrastructure. By targeting established industries, European policymakers are attempting to engineer an economic transition rather than merely chasing Silicon Valley’s consumer-focused tail.

That said, selling applied intelligence requires an entirely different sales motion. Startups must now prove integration capabilities with legacy SAP and Oracle databases. They have to navigate complex procurement cycles. The romantic era of the overnight AI unicorn is dead. We are now in the era of the gruelling enterprise sales cycle, where security audits matter more than parameter counts.

This transition toward utility is not happening in a vacuum. It is being heavily engineered by Brussels. The enforcement of the European AI Act has fundamentally altered the structural economics of software development on the continent. Critics initially warned that the legislation would stifle innovation, but the reality on the ground at VivaTech suggests a different outcome. Regulation has inadvertently created a massive market for compliance-grade, sovereign AI solutions.

What are the main AI trends at VivaTech?

At VivaTech, the primary AI trends centre on applied artificial intelligence, strict regulatory compliance under the EU AI Act, and enterprise-grade deployment. Companies are actively abandoning generative novelty in favour of measurable productivity gains, secure sovereign data solutions, and demonstrable return on investment.

This compliance-first approach offers a distinct competitive moat. American tech giants are currently battling copyright infringement lawsuits and regulatory scrutiny regarding their data scraping methodologies. European startups, conversely, are building models explicitly trained on licensed, opt-in data. They are offering guarantees that foreign competitors cannot match. When a German automotive manufacturer integrates a predictive maintenance model, they require absolute certainty that their proprietary telematics data will not be used to train a public model.

The picture is more complicated than a simple trans-Atlantic rivalry. It is a divergence in product philosophy. The US model prioritises general intelligence and rapid consumer adoption. The emerging European model, showcased vividly across the VivaTech pavilions, prioritises domain-specific accuracy, data sovereignty, and legal safety. In the enterprise sector, safety is rapidly becoming a premium feature rather than a bureaucratic burden.

The downstream consequences of this shift are profound for both policymakers and small-to-medium enterprises (SMEs). For the latter, the barriers to entry are finally lowering. For the last three years, AI deployment was effectively restricted to multinational corporations with vast engineering resources. The current generation of applied tools, heavily promoted at VivaTech, operates as plug-and-play software.

This democratization of capability will aggressively disrupt traditional B2B service sectors. Legal research, entry-level accounting, and supply chain logistics are facing immediate margin compression. According to a recent analysis by Bloomberg Intelligence, professional services firms that fail to adopt automated workflows will see their operating margins contract by up to 15% over the next 24 months. The cost of remaining analogue is becoming fatal.

Still, this transition requires massive infrastructure. The bottleneck has shifted from software capability to physical compute. Sovereign data solutions demand localized data centres. European nations are currently scrambling to build the requisite energy and cooling infrastructure to support this localized compute demand. The next major geopolitical battleground will not be the algorithms themselves, but the raw gigawatts required to run them domestically.

Governments are acutely aware of this vulnerability. French President Emmanuel Macron used his opening address at the conference to announce accelerated permitting processes for green-energy data centres. The goal is clear: to ensure that the intellectual property generated by European applied AI remains physically housed within the borders of the European Union.

Competing Perspectives: The Compute Deficit and Market Fragmentation

Not everyone in the halls of Porte de Versailles shares this optimistic vision of a European industrial renaissance. A vocal contingent of investors argues that the continent’s focus on applied AI is essentially a concession of defeat in the foundational model race. The bear case is structural and compelling.

Europe remains fragmented. A startup cannot scale across the continent without navigating 27 different legal jurisdictions and language barriers. More critically, the hardware deficit is severe. According to Reuters technology analysts, Europe currently accounts for less than 12% of the global advanced GPU supply. You cannot build a sovereign AI ecosystem if you rely entirely on Californian hardware manufactured in Taiwan.

Dissenting voices argue that by focusing purely on B2B applications, European firms risk becoming entirely dependent on the foundational API layers controlled by OpenAI, Google, and Anthropic. If the base cost of inference rises, the profit margins of these European applied AI companies will collapse. In this view, the regulatory moat created by the AI Act is a temporary illusion, easily breached once the foundational models reach a threshold of undeniable superiority.

Yet, the counter-argument remains potent. Foundational models are rapidly commoditising. Open-source alternatives are narrowing the performance gap weekly. If intelligence becomes a cheap, ubiquitous utility, the real economic value will accrue to the companies that own the proprietary workflow integrations and the industry-specific data.

Ten years on, VivaTech has shed its adolescent idealism. The focus on artificial intelligence that practically functions within the rigid constraints of modern business represents a necessary maturation. Europe is no longer attempting to clone Silicon Valley. It is building an ecosystem tailored to its own industrial and regulatory DNA.

The tension between foundational dependence and applied utility will define the next decade of enterprise technology. However, the mood in Paris suggests a quiet confidence that the pendulum is swinging back toward business fundamentals. The era of the speculative demo has officially concluded; the era of ruthless execution has begun.

The Week Wall Street Ran Two Separate Races

On Monday, May 11, three of America’s most-watched indices — the S&P 500, the Nasdaq Composite, and the Dow Jones Industrial Average — closed simultaneously at record highs. By Friday, the party was over for tech, with Nvidia shedding 4.4% and Intel retreating more than six percent in a single session, as Treasury yields spiked and traders remembered that gravity is still a law. Yet even in that churn, a clear list of winners emerged: companies levered to artificial intelligence infrastructure, geopolitically sensitive energy, and a rearming defence sector. Here are the ten US stocks that mattered most this week — and why.

Context: A Market at an Altitude It’s Never Seen Before

The S&P 500 achieved its seventh consecutive weekly gain as of May 11, with the index sitting at 7,412.84. Information Technology, Communication Services, and Consumer Discretionary led sector performance, while the rally was notably narrow — the equal-weight S&P significantly underperformed its cap-weighted counterpart, pointing to concentration in a handful of mega-cap names. Tradingkey

Underneath that headline number, the macro picture is genuinely complicated. First-quarter 2026 real GDP grew at an annualised rate of 2.0%, driven primarily by business investment in AI-related equipment and software, while consumer spending grew at a slower 1.6% pace. The Federal Open Market Committee held the federal funds rate steady at 3.5% to 3.75% at its April meeting, even as Jerome Powell concluded his tenure on May 15 and Kevin Warsh took over as Fed Chair. Oil is the wild card in the room: Brent crude surged 2.9% to above $104 per barrel on May 11 after President Trump described the US-Iran ceasefire as “on life support,” rekindling inflation fears. Tradingkey + 2

The market, in other words, is running two separate races. One is the AI infrastructure buildout, where capital expenditure is still accelerating. The other is a geopolitical energy trade that is increasingly testing consumer resilience. The ten stocks below sit at the intersection of both.

The Top 10 US Stocks Profitable This Week

These are not predictions. They are a snapshot of where market energy, earnings momentum, and institutional conviction converged during the week of May 12–19, 2026.

1. Rackspace Technology (RXT)

The week’s most dramatic story belongs to a company that was written off as a legacy data-centre casualty two years ago. Rackspace Technology surged over 165% in May on the back of hyperscaler partnerships and AI infrastructure capacity expansion, with strong Q1 results and an upgraded full-year outlook triggering a wave of short-covering and institutional buying. Analysts have upgraded the stock to Buy with price targets above $15. It’s a small-cap proxy on the same AI infrastructure theme powering the giants — but with the volatility that comes with a fraction of their market cap. Tradingkey

2. Nvidia (NVDA)

Nvidia reached its all-time high of $236.54 on May 14, 2026, with a market capitalisation of $5.46 trillion as of this week. Every number that matters is pointed upward. In fiscal year 2026, Nvidia’s revenue hit $215.94 billion — a 65.47% increase year-on-year — with earnings of $120.07 billion. The company reports Q1 fiscal 2027 results on May 20. What Jensen Huang says about the forward demand picture may matter more than the print itself. TradingViewStockAnalysis

3. Alphabet (GOOGL)

Alphabet has been the single biggest engine of the S&P 500’s 2026 rally, contributing 1.27 percentage points to the index’s return — more than 20% of the index’s total gain from one name alone. Google Cloud demand is accelerating, Gemini is gaining traction in the enterprise market, and the market is finally giving Alphabet credit for its custom AI chips — TPUs — as a credible alternative to Nvidia’s GPUs. The stock recently leapfrogged Apple for the number two spot by global market capitalisation. ETF.com

4. Arista Networks (ANET)

Arista reported Q1 2026 revenue of $2.71 billion against a consensus of $2.62 billion, representing 35% year-on-year revenue growth, with net income rising to $1.02 billion from $813.8 million. The company raised its full-year 2026 revenue guidance to $11.5 billion. Microsoft, Meta, Alphabet, and Amazon have guided combined capital expenditure above $320 billion for 2026, and every dollar of that spend on GPU clusters eventually flows through the ethernet switching market that Arista dominates. StockAnalysisGotrade

5. Broadcom (AVGO)

Broadcom sits second only to Alphabet in its contribution to the S&P 500’s 2026 gains, adding 0.6 percentage points from an average index weight of just 2.8%. Its custom AI silicon partnerships with Google, Meta, and other hyperscalers give it a structural position in the AI supply chain that is less visible than Nvidia’s but no less valuable. ETF.com

6. Innodata (INOD)

Innodata posted triple-digit gains in May on the back of AI data annotation contracts with large-language-model developers. It’s a pick-and-shovel play on the one input that every AI model needs before it can generate a single token: high-quality labelled training data. With frontier model labs locked in an arms race, demand for that service isn’t slowing. Tradingkey

7. Fluence Energy (FLNC)

Fluence Energy soared close to 30% in the week after HSBC and Roth Capital both upgraded the stock following fiscal second-quarter EBITDA that topped Wall Street estimates — the stock had already rocketed roughly 40% the prior session. AI data centres are power-hungry at a scale that demands grid-scale battery storage solutions. Fluence, which sells exactly that, is riding the intersection of energy demand and AI infrastructure. CNBC

8. Lockheed Martin (LMT)

Lockheed Martin was among the week’s gainers as renewed US-Iran tension kept WTI crude near $105 per barrel, with markets pricing in increased Pentagon outlays for Middle East uncertainty and sustained great-power competition with China. The company announced a quarterly cash dividend of $3.45 per share with an ex-date of June 1. In a week where growth stocks slid on Friday, LMT offered something that few technology names can: a reason to hold that doesn’t depend on the next earnings beat. Tradingkey

9. RTX Corporation (RTX)

The same geopolitical current lifted RTX. The energy sector was the sole sector to post gains on Friday, May 15, rising 1.6%, while defence names including RTX benefited from the market pricing in higher Pentagon spending tied to Middle East friction and the broader US military posture. RTX’s exposure to both the missile stockpile-replenishment cycle and the commercial aerospace aftermarket gives it two separate earnings engines — a rare structural advantage in an uncertain macro environment. Tradingkey

10. P3 Health Partners (PIII)

The month’s most extreme mover. P3 Health Partners posted the highest monthly gain of any NYSE or Nasdaq stock in May 2026, with a rise of 285%. The managed-care company’s surge is event-driven, tied to Medicare Advantage contract developments and a reassessment of its financial trajectory. It is also exactly the kind of move that attracts momentum traders, which can amplify both the upside and the eventual correction. Stocktitan

The Structural Story Behind the Numbers: Why Are These Stocks Really Moving?

Is the AI stock rally sustainable heading into the second half of 2026?

The AI rally’s staying power ultimately rests on one question: are the hyperscalers getting returns on their capital expenditure, or are they building infrastructure that will take years to monetise? The evidence so far favours optimism — cautiously. With approximately 89% of S&P 500 companies having reported Q1 results, the index showed year-on-year revenue growth of 10.4% and earnings growth of 25.3%. Those are not the numbers of a market hallucinating its own prosperity. Tradingkey

Yet the rally’s narrowness is a legitimate concern. When Alphabet alone accounts for more than a fifth of the S&P 500’s total 2026 return, portfolio concentration has moved from a feature to a risk. The market’s gains have been described by analysts as narrow, with the equal-weight S&P significantly underperforming its cap-weighted version — a sign that broader market participation has not kept pace with mega-cap appreciation. CNBC

Featured snippet answer — What are the top performing US stocks this week? The top performing US stocks for the week of May 12–19, 2026 include Rackspace Technology (RXT), Nvidia (NVDA), Alphabet (GOOGL), Arista Networks (ANET), Broadcom (AVGO), Innodata (INOD), Fluence Energy (FLNC), Lockheed Martin (LMT), RTX Corporation (RTX), and P3 Health Partners (PIII). Their gains are driven by AI infrastructure demand, rising defence spending, and geopolitical oil premiums from the ongoing Iran conflict.

The second structural driver — energy and defence — is less discussed but may prove stickier. The Strait of Hormuz carries roughly 20% of global oil and LNG supply, and geopolitical scenarios around the US-Iran ceasefire have become materially priced into markets, with WTI trading near $105 per barrel. That’s not a trade; it’s a repricing of geopolitical risk that could persist for months. Tradingkey

Implications and Second-Order Effects

The week’s price action carries downstream consequences that go well beyond the tick-by-tick narrative.

First, Nvidia’s May 20 earnings report will function as a referendum on the entire AI supply chain. Consensus estimates for the report point to continued data centre revenue growth exceeding 60%, and a beat-and-raise result would likely sustain the infrastructure buildout trade across chips, networking, and cloud computing names. A miss, or a conservative guide on data centre demand, would reprice not just NVDA but Arista, Broadcom, and the broader semiconductor ecosystem simultaneously. As the TradingKey analysis put it bluntly: every AI trade next week is binary to that print. Tradingkey

Second, the spike in 30-year Treasury yields — which jumped above 5.1% on Friday, May 15, the highest since May 2025 — introduces a genuine valuation headwind for long-duration growth assets. Higher yields compress the present value of future earnings. For companies like Arista and Broadcom, whose valuations embed years of high-growth assumptions, that compression isn’t trivial. The bond market, in other words, is not convinced that the AI story justifies current multiples. CNBC

Third, the energy premium from the Iran situation is starting to attract the attention of recession forecasters. Dan Niles, founder of Niles Investment Management, told CNBC on May 15 that ten of the last twelve recessions were preceded by an oil price spike, and that the Federal Reserve’s ability to cut rates could be compromised by oil’s inflationary effect. Traders now lean toward rate hikes as the Fed’s next move — a reversal of expectations that would represent a significant tightening of financial conditions for the consumer. CNBC

For investors in defence stocks like LMT and RTX, the implications are more favourable. Pentagon budgets tend to expand under geopolitical pressure regardless of the broader economic cycle, and the current administration’s posture toward both Iran and China suggests a multi-year tailwind that doesn’t depend on any single quarter’s earnings surprise.

The Bear Case Deserves a Hearing

Not everyone is reading the rally’s signal the same way.

Michael Burry drew attention this week by comparing the Philadelphia Semiconductor Index’s trajectory — up more than 10% in a single week, with 2026 gains reaching 65% — to the run-up that preceded the technology collapse of March 2000. The comparison is inexact: the current semiconductor cycle is underpinned by real revenue growth rather than projected eyeballs. Still, the pace of the move has concentrated enough wealth in a narrow band of names to make a reversal systemically significant. CNBC

The sceptics also point to the rally’s engine. Alphabet’s outsized contribution to S&P 500 returns is, structurally, the same problem the index had in 2020–21 with a different name at the top. Single-name concentration at the index level means passive investors are more exposed to Alphabet’s fortunes than they may realise — and more exposed to any negative development in the EU’s regulatory approach to Google’s AI integration or its search dominance.

There’s a third concern: the retail investor sentiment data suggests that individual traders have been buying heavily into the top momentum names. The SPDR S&P Retail ETF fell more than 6% across the week of May 12–16, its fourth consecutive weekly decline, as investors grew cautious on the consumer backdrop and discretionary spending. A divergence between the market that Wall Street trades and the economy that Main Street inhabits is not indefinitely sustainable. CNBC

That said, earnings remain the ultimate arbiter. With year-on-year earnings growth across the S&P 500 running at 25.3%, the fundamental case for current valuations is more defensible today than it was in early 2000 — when many of the index’s leaders were pre-revenue concepts dressed up as infrastructure plays. Tradingkey

Closing

The ten stocks that led the market this week are not a random collection of fortunate names. They are a map of where capital is flowing in 2026: into the infrastructure of artificial intelligence, into the energy markets shaped by geopolitical fracture, and into the defence complex of an America that is visibly rearming. Whether that map remains accurate depends on what Nvidia says Wednesday evening, what Kevin Warsh signals about the rate path, and whether WTI can stay above $100 without breaking the consumer who ultimately funds all of it.

The week offered a sharp reminder that the best-performing stocks are rarely the whole story. The energy sole sector that rose on Friday while technology fell was not a coincidence. It was a rotation — provisional, perhaps, but pointed. In markets running at this altitude, what leads one week can lag the next. The investors who’ll do well in the second half of 2026 won’t be the ones who bought the top of the momentum list. They’ll be the ones who understood why each stock was on it.

The rally is still alive. The questions are just getting harder.