Five deals defined the week ending May 28: a $7 billion cancer biotech buyout that may set a new venture-backed record, a €11.5 billion takeover battle that sent an online food-delivery giant's stock to an 18-month high, a $1.3 billion royalty financing that quietly restructured how late-stage biotech cash flows work, Snowflake doubling down on cloud infrastructure with a $6 billion AWS commitment, and a $225 million data-center bet on AI inference. Here is the full picture.

Acquirer: Eli Lilly | Target: Kelonia Therapeutics | Value: up to $7 billion cash ($3.25 billion upfront + up to $3.75 billion in milestones)

Lilly moved fast on in-vivo cell therapy this week, announcing the acquisition of Boston-based Kelonia Therapeutics in what Crunchbase called the largest acquisition of a venture-backed private biotech in roughly a decade. Kelonia has operated for only four years, emerging from stealth with a $50 million Series A led by Alta Partners, Horizons Ventures and Venrock. Its core technology lets doctors reprogram a patient's T cells directly inside the body — no extraction, no lab processing — to hunt and kill tumour cells. The approach is sometimes called "off-the-shelf" in-vivo CAR-T, and early clinical data in multiple myeloma was strong enough to attract Lilly's biggest deal of the year.

Strategic intent: Lilly is building an in-vivo cell therapy franchise at speed. Two months before the Kelonia deal it paid up to $2.4 billion for Orna Therapeutics, another in-vivo-CAR-T player. The pattern is clear: rather than wait for Phase 3 readouts, Lilly is buying platform technologies early, absorbing milestone risk in exchange for owning the IP before competitors can get in.

Financier: Blackstone Life Sciences | Beneficiary: Apogee Therapeutics | Structure: up to $800 million royalty financing + up to $500 million senior corporate debt

This one is not a straight acquisition, but it reshapes Apogee's competitive posture as sharply as any buyout would. Blackstone Life Sciences agreed to provide up to $800 million in exchange for low-to-mid single-digit tiered royalties on zumilokibart — Apogee's anti-IL-13 antibody for moderate-to-severe eczema — for 15 years. An additional $500 million in senior debt is available subject to mutual agreement, bringing the headline to $1.3 billion.

Apogee's eczema drug already looked like a contender: Phase 2 mid-stage data showed 65.9% of patients at the mid-dose hitting a 75%-or-greater skin clearance at 16 weeks, putting it in the same tier as Lilly's Ebglyss and Sanofi/Regeneron's Dupixent on cross-trial comparisons. Taking it into pivotal trials requires capital beyond what Apogee had on its balance sheet ($1.3 billion at end of Q1), hence the non-dilutive royalty structure.

Strategic intent: Blackstone frames this as a template — "largest royalty financing for a pre-Phase 3 program to date." For BXLS, which also backed Anagram Therapeutics ($250 million), Teva ($400 million) and Merck ($700 million) in recent quarters, royalty financing is a structured way to get drug-asset exposure without full acquisition risk. For Apogee, it preserves equity while funding a clear path to commercialisation.

Acquirer: Uber | Target: Delivery Hero (DHER.DE) | Bid value: €11.5 billion (€33/share, initial offer rejected; board reviewing raised bid)

The week's most dramatic deal story: Uber disclosed on May 22 (Bloomberg) that it was exploring a full acquisition of Delivery Hero, the Berlin-based online food-delivery group that operates Foodpanda, Talabat and other brands across 40+ countries. On May 23, Delivery Hero confirmed it had received Uber's formal proposal at €33 per share, implying a total enterprise value above €11.5 billion (~$13.4 billion). Major shareholders rejected that price. By May 24 Uber's board had convened to discuss raising the offer. By May 27, regulatory filings showed Uber had lifted its stake from 19.5% to 24.99% — the maximum a single shareholder can hold in Germany before triggering a mandatory full bid.

Strategic intent: Uber has been quietly circling Delivery Hero for months. It already owned a near-20% stake from a 2022 asset swap. A full deal would give Uber a single global delivery infrastructure spanning markets where it has weak presence — particularly Southeast Asia (Foodpanda) and the Middle East (Talabat). In parallel, Uber and South Korea's Naver tabled a bid of up to ₩8 trillion (~$5.4 billion) for Baedal Minjok (Baemin), Delivery Hero's Korean subsidiary, signalling that Uber may pursue a break-up of the group rather than a unified acquisition.

Parties: Snowflake and Amazon Web Services | Value: $6 billion over five years | Core asset: AWS Graviton ARM-based CPU chips

Snowflake and AWS announced a new $6 billion, five-year infrastructure agreement, centred on AWS's in-house Graviton ARM chips. For context: Snowflake has transacted approximately $7 billion in total lifetime revenue via the AWS Marketplace since its 2012 founding — so this single new contract is nearly that entire historical figure.

The driver is agentic AI. As enterprises run AI agents continuously rather than in occasional inference bursts, CPU demand balloons relative to GPU. Snowflake's Cortex AI (its text-to-SQL and agent-orchestration product) sits on top of petabytes of customer data that already live on Snowflake/AWS. Locking in Graviton supply gives Snowflake preferential pricing for the compute surge ahead.

Strategic intent: Snowflake raised its full-year product revenue guidance at the same time as the AWS announcement, signalling that the deal is not defensive but tied to visible demand growth. AWS customer spending on Snowflake doubled in 2025 alone, to $2 billion for that calendar year. For AWS, the deal is a counter-move to Meta's recent $10 billion Google Cloud commitment — showing that Graviton can win hyperscaler-scale deals.

Acquirer: I Squared Capital | Target: Cogent Communications data center | Value: $225 million | Rationale: AI inference positioning

I Squared Capital, the global infrastructure private-equity firm, agreed to pay $225 million to acquire a data center from Cogent Communications, the internet backbone and colocation provider. The deal was flagged on Axios Pro Deals (May 26) with an explicit AI inference rationale: I Squared is building out compute infrastructure to serve enterprise and hyperscaler demand for AI inference workloads, which require low-latency, high-density colocation rather than the large campus GPU clusters used for training.

Strategic intent: I Squared's playbook here mirrors broader PE behaviour in 2026 — buy physical compute real estate before the build cycle peaks, then lease capacity to AI model operators and enterprises that cannot wait for greenfield construction timelines. Cogent, historically a network-focused business, is monetising non-core real estate while retaining network relationships.

In-vivo cell therapy is the new battleground. Lilly's Kelonia deal follows a pattern visible across big pharma: rather than licensing in-vivo CAR-T programs, acquirers are buying the platform companies outright while they are still pre-Phase 3. Expect Bristol Myers Squibb and Pfizer to respond.

Royalty financing as an M&A alternative. Blackstone's Apogee deal illustrates a maturing market for non-dilutive late-stage biotech funding. As IPO windows stay narrow, royalty structures let cash-hungry programs avoid both equity dilution and a full-sale price tag, while giving financial acquirers drug-asset exposure.

Uber's logistics empire ambitions. The Delivery Hero move is the clearest sign yet that Uber views itself not as a ride-share company with a food delivery sideline, but as a global logistics platform. If Uber closes the Delivery Hero deal and simultaneously acquires Baemin, it would overnight become the dominant delivery operator across Southeast Asia, the Middle East, and parts of Europe.