Oil prices edged higher on Friday, but the market remained on track for its first weekly decline since early February, as traders balanced tentative signs of de-escalation against the deeper reality of a prolonged supply shock tied to the Iran conflict.

Brent crude rose $1.87, or 1.73%, to $109.88 a barrel, while U.S. West Texas Intermediate (WTI) gained $1.57, or 1.66%, to $96.05. The gains offered only partial relief after a week of consolidation, with Brent down 2.1% and WTI off 2.3% over the period.

The weekly dip follows an extraordinary run-up. Since the outbreak of hostilities in late February, Brent has surged more than 50%, while WTI has climbed over 40%, reflecting a market that has aggressively priced in geopolitical risk and the threat of sustained supply disruption.

Attention remains fixed on policy signals from Donald Trump, whose decision to extend a pause in attacks on Iran’s energy infrastructure has cooled immediate fears of further escalation. That restraint has given traders an opportunity to take profits, but few are interpreting it as a definitive shift toward de-escalation.

Instead, the market is increasingly trading on the assumption that the conflict will endure. Washington has set an April 6 deadline for Iran to reopen the Strait of Hormuz, one of the world’s most critical oil transit routes, while simultaneously reinforcing its military presence in the region. The possibility of direct action against key Iranian export infrastructure, including Kharg Island, continues to hang over the market.

The scale of disruption already priced into crude is substantial. Roughly 11 million barrels per day have been taken out of global supply, according to industry estimates — a shock the International Energy Agency has characterized as more severe than the oil crises of the 1970s.

Flows through the Strait of Hormuz remain constrained, effectively removing a significant portion of Middle Eastern exports from the market. That tightening is beginning to ripple through physical markets, with refiners facing higher input costs and longer delivery times, particularly in Asia, where dependence on Gulf crude is most acute.

“Every additional day of restricted flows compounds the deficit,” UBS analyst Giovanni Staunovo noted, highlighting that more than 10 million barrels per day remain disrupted.

The result is a market that is tightening not just on paper but in physical availability, with spot premiums widening in key trading hubs.

The response from consuming nations is becoming more visible. Strategic reserves are being tapped, particularly across Asia, while some economies are beginning to adjust demand expectations. Mukesh Sahdev of consultancy XAnalysts said the longer the conflict persists, the greater the likelihood of structural demand responses, including reduced industrial consumption and shifts toward alternative fuels.

At the same time, the market is grappling with divergent issues. Analysts at Macquarie Group argue that a near-term easing of hostilities could trigger a sharp correction in prices, though not a full return to pre-conflict levels, given residual risk. In contrast, a prolonged conflict stretching into mid-year could push crude toward $200 a barrel — a level that would amplify inflationary pressures globally and complicate central bank policy.

The broader macroeconomic implications are already coming into focus. Elevated oil prices risk feeding into transport and manufacturing costs, potentially reversing recent gains in inflation control across major economies. For emerging markets, particularly energy importers, the shock threatens to widen trade deficits and put pressure on currencies.

Yet for all the volatility, the market’s underlying posture remains defensive. Traders are unwinding some positions after a steep rally, but few are willing to step away entirely, given the asymmetry of risk — where downside appears limited in the short term, but upside could be abrupt and severe if the conflict escalates.

The weekly decline is currently seen as a pause driven by tactical repositioning rather than a shift in fundamentals. The dominant forces shaping the market, constrained supply, geopolitical uncertainty, and fragile transit routes, remain firmly in place. In that context, oil is no longer reacting solely to events on the ground, but to the duration and potential expansion of the conflict itself.

The Nasdaq officially has entered correction territory Thursday, tumbling more than 2 percent as the S&P 500 and Dow Jones each dropped over 1 percent.

Investors dumped risk assets amid a thickening “fog of war” in the Middle East, where a month-old U.S.-Israeli campaign against Iran shows no clear path to resolution despite President Donald Trump’s announcement of a 10-day pause in strikes on Iranian energy facilities.

The selling wave marked the steepest one-day loss for the Nasdaq and S&P 500 since January 20. By the close, the Dow had shed 469.38 points, or 1.01 percent, to finish at 45,960.11. The S&P 500 gave up 114.74 points, or 1.74 percent, closing at 6,477.16. The Nasdaq Composite plunged 521.74 points, or 2.38 percent, to 21,408.08 — now 10.7 percent below its October 29 record high.

Trump’s late-day statement that talks with Tehran were “going very well,” and that attacks on energy plants would halt until April 6 at Iran’s request, helped stock futures trim losses after the bell. But the damage was already baked in. Earlier in the session, the absence of any tangible diplomatic breakthrough, coupled with Tehran’s dismissal of the U.S. proposal as “one-sided and unfair”, sent oil prices rocketing higher. U.S. crude settled up 4.6 percent; Brent jumped 5.7 percent.

The reason is straightforward and ominous: roughly 20 million barrels a day of crude and refined products, about one-fifth of global oil consumption and a quarter of all seaborne trade, normally flow through the narrow Strait of Hormuz. With shipping already disrupted, any prolonged closure or threat of closure turns a regional conflict into a global inflation shock.

Doug Beath, global equity strategist at Wells Fargo Investment Institute, captured the mood perfectly, noting: “The back and forth seems to be happening at a quicker pace. On top of it, we don’t know who Trump is negotiating with. There’s a lot of conflicting signals, and it’s really the fog of war, the uncertainty of all of it that’s driving this.”

With no solution in sight after four weeks of fighting, markets are now openly gearing up for the worst. The OECD warned Thursday that the conflict has already derailed the global economy from a stronger growth track, with near-term risks of sharply higher inflation if Hormuz flows remain throttled. Central banks, already navigating sticky prices, now face a classic policy trap: higher energy costs feeding into broader inflation just as growth momentum fades.

Traders have scrubbed any expectation of Federal Reserve rate cuts this year; two had been priced in before the bombs started falling.

The sell-off carried a familiar post-pandemic flavor but with a sharper geopolitical edge. After three straight years of strong gains, powered largely by the AI-fueled tech rally, a 10-to-20 percent pullback “should not surprise anyone,” said Peter Tuz, president of Chase Investment Counsel.

“We had one last year during the tariff proposals. Bad technical indicators might, however, encourage selling and discourage buying until the situation clears up,” he said.

Most S&P 500 sectors finished in the red. Energy was the clear winner, up 1.6 percent as investors sought shelter in the very commodity that was inflating costs elsewhere. Defensive utilities managed a modest 0.2 percent gain. The heaviest beatings came in communications services, down 3.5 percent, and technology, off 2.7 percent.

Chip stocks led the tech rout. The Philadelphia Semiconductor Index cratered 4.8 percent after three days of tentative gains. Nvidia, the face of the artificial-intelligence boom, fell more than 4 percent as higher energy prices threaten the power-hungry data centers that train the next generation of models. Communications services took a separate hit after a Los Angeles jury held Meta and Alphabet’s Google liable in the first wave of lawsuits accusing social-media platforms of harming children. Meta shares closed nearly 8 percent lower; Alphabet lost more than 3 percent.

Gold prices slipped more than 2 percent, dragging down gold-miner stocks such as Sibanye Stillwater and Harmony Gold by over 4 percent each. Even the traditional safe haven felt the pull of fragile hopes that Trump’s pause might stick.

Trading volume stayed light, just 16.5 billion shares across U.S. exchanges versus a recent 20-day average of 20.5 billion, a classic sign of nervous hesitation rather than outright panic. Decliners swamped advancers roughly 3-to-1 on the NYSE and 2.5-to-1 on the Nasdaq.

The broader picture is one of exhausted optimism. What began as a calculated military strike has morphed into an open-ended economic drag. Higher oil not only fans inflation; it squeezes corporate margins, crimps consumer spending, and raises the specter of “demand destruction” in energy-intensive industries.

The VIX, Wall Street’s fear gauge, has remained elevated above 20 since the conflict erupted and spiked as high as 35 in the early days — levels that signal investors are pricing in more volatility ahead, not less.

Anthropic has begun quietly testing a new frontier AI model that it describes as a clear “step change” beyond anything it has released before.

The company acknowledged Thursday after draft documents detailing the project were accidentally left exposed in a public data cache, according to Fortune.

The model, internally referred to as both Claude Mythos and Capybara, would introduce an entirely new tier above the company’s current flagship Opus line. According to the leaked draft blog post reviewed by Fortune, Capybara is “larger and more intelligent than our Opus models — which were, until now, our most powerful.”

It delivers dramatically higher performance on benchmarks for software coding, academic reasoning, and especially cybersecurity tasks compared with Claude Opus 4.6. Anthropic spokesperson confirmed the company is developing “a general purpose model with meaningful advances in reasoning, coding, and cybersecurity.”

The spokesperson added: “Given the strength of its capabilities, we’re being deliberate about how we release it… We consider this model a step change and the most capable we’ve built to date.”

The documents surfaced through a straightforward configuration error in Anthropic’s content management system. Assets uploaded to the CMS were set to public by default, leaving nearly 3,000 unpublished files, including images, PDFs, audio, and the draft announcement, searchable and downloadable by anyone.

Cybersecurity researchers Roy Paz of LayerX Security and Alexandre Pauwels of the University of Cambridge spotted the cache and alerted Fortune. Once notified on Thursday, Anthropic quickly locked down public access.

The company attributed the lapse to “human error” in configuring an external CMS tool and described the exposed material as early drafts considered for publication.

A Cautious Rollout and Major Cyber Concerns

Anthropic is taking an unusually measured approach to the launch. The model is currently in early-access trials with a small group of customers, and the draft makes clear it is too expensive and potentially too risky for immediate general release.

The biggest red flag highlighted in the leaked document is cybersecurity. Anthropic warns that the model is “currently far ahead of any other AI model in cyber capabilities” and “presages an upcoming wave of models that can exploit vulnerabilities in ways that far outpace the efforts of defenders.”

Hackers armed with such a system could launch large-scale, automated attacks on codebases at a speed and sophistication that current defenses may struggle to match.

Because of that risk, the company’s plan emphasizes giving cyber defenders a head start.

“We’re releasing it in early access to organizations, giving them a head start in improving the robustness of their codebases against the impending wave of AI-driven exploits,” the draft stated.

This mirrors emerging incidents of the growing industry. In February, OpenAI flagged its GPT-5.3-Codex as the first model it classified as “high capability” for cybersecurity under its Preparedness Framework. Anthropic’s own Opus 4.6, released around the same time, already showed strong dual-use potential — capable of surfacing unknown vulnerabilities in live code, a tool that could help attackers as easily as it helps defenders.

The company has also documented real-world attempts by Chinese state-linked groups to weaponize earlier Claude versions for coordinated intrusions into tech firms, banks, and government agencies.

New Model Tier and Enterprise PushThe leak also revealed Anthropic’s intention to reshape its product lineup. Until now, Claude models have come in three sizes: Haiku (fast and cheap), Sonnet (balanced), and Opus (most capable). Capybara would sit above Opus as a premium, higher-cost tier — larger, smarter, and significantly more expensive to run.

The documents further exposed plans for an invite-only, two-day executive retreat in the English countryside. Scheduled at an 18th-century manor turned luxury hotel and spa, the gathering is aimed at Europe’s most influential CEOs. Anthropic CEO Dario Amodei is expected to attend, and participants will hear from policymakers on AI adoption while getting hands-on exposure to unreleased Claude capabilities.

The company described the event as part of an ongoing series to court large corporate customers.

Anthropic confirmed the retreat is real and fits its broader strategy of deepening relationships with enterprise leaders.

The development underscores the high-stakes environment in which frontier AI labs now operate. Even a simple misconfiguration in a routine content system can spill sensitive product details, internal strategy, and risk assessments into the open. For a company that has positioned itself as the more safety-conscious alternative to OpenAI, the leak is an unwelcome reminder that operational hygiene matters as much as model alignment when capabilities reach this level.

Anthropic has not set a public release date for the new model, saying only that it will move deliberately. In the meantime, the early-access program will likely serve as both a testing ground and a controlled way to let trusted partners begin hardening their systems against the next wave of AI-powered cyber threats.

The incident comes as competition among the leading labs intensifies, with each new model promising bigger leaps and bigger headaches, in capabilities that blur the line between powerful tool and potential weapon. Anthropic’s challenge now is to prove it can handle the power it is building while keeping its own house in order.