Shares of Tempus AI (NASDAQ: TEM) tumbled 10% after the company was targeted in a short-seller report by Spruce Point. The report raised serious concerns about the integrity of Tempus AI’s product, the credibility of its management, and its financial reporting practices.

The Spruce Point report, which follows a forensic review of the precision medicine solutions company, criticized Tempus AI for the minimal revenue generated from its AI Applications product line, suggesting a disconnect between the company’s branding and its financial reality. In 2024, AI-related revenue was reported to be just $12.4 million, less than 2% of the company’s total revenue of $693.4 million. The prior year, AI revenue was only $5.5 million.

Moreover, the report highlighted concerns over Tempus AI’s leadership, noting CEO Eric Lefkofsky’s history with companies that experienced financial scandals, including bankruptcies and restatements. It also questioned the company’s partnerships and joint ventures, suggesting potential round-tripped sales and related party deals.

The report further scrutinized Tempus AI’s billing practices, indicating aggressive use of billing code 81479, known for minimal oversight and a high risk of abuse. This tactic could inflate revenue at the expense of patient care, with the report suggesting that Tempus AI and its acquired entity, Ambry, may prioritize revenue over clinical necessity.

In conclusion, Spruce Point warned investors of the company’s reliance on perceived equity value, aggressive accounting, and unsustainable practices. They projected a 50% – 60% potential downside risk to Tempus AI’s share price, advising extreme caution and close scrutiny of the company’s public statements. The sell-side consensus price target stands at $66.82 per share, which implies a mere 1% upside potential.

Spruce Point Report Overview

Tempus AI, Inc., (formerly Tempus Labs, Inc.) based in Chicago, IL, provides AI-enabled precision medicine solutions. The Company develops its diagnostics through the application of artificial intelligence in healthcare to make laboratory tests and connect lab results to a patient’s clinical data. As of the last twelve months ended March 31, 2025, the Company reported approximately $803 million and ($77) million of revenue and Adjusted EBITDA, respectively.

The concerns we outline in our report include:

• Tempus Founder Eric Lefkofsky and his associates have a history of promoting disruptive technology companies, cashing out early, and leaving public shareholders with losses or lackluster returns.

  • Mr. Lefkofsky and his associate Bradley Keywell, who is also a major Tempus shareholder, have a history of partnering on companies that have ended poorly for public investors. The partners first sold Starbelly.com to HA-LO Industries during the dot-com bubble which was blamed by some for sending HA-LO into bankruptcy. Starbelly.com was positioned as the next Dell.com but that concept never came to fruition and litigation emerged.

  • The partners later co-founded Echo Global Logistics and Groupon.

    • Groupon’s struggles have been widely publicized, and despite Mr. Lefkofsky’s claims that Groupon would be “wildly profitable” and that some of its offerings could be positioned against Costco, its success was short-lived and its share price has declined -95% since its IPO. During the IPO process, Groupon also restated revenues by nearly 50% after encountering resistance from the SEC about its revenue recognition policy, and it reported a material weakness of internal controls after the IPO.

    • Echo Global Logistics was touted as a technology disruptor in the freight and logistics market and also reported a material weakness of internal controls after its IPO. While public investors were ultimately rewarded with a takeover premium of +244% twelve years after its IPO, the stock vastly underperformed the Dow Jones Transportation Index by approximately 110% and the takeover came years after the partners ceased active involvement.
      

• Concerns around Tempus’ AI capabilities are heightened given the minimal revenue generated from its AI applications.

  • Amid the market enthusiasm for artificial intelligence, Tempus Labs rebranded as Tempus AI before its IPO. Despite its name, Tempus generated only $12.4 million, or 2% of total revenues, from AI applications in 2024. This time around, Mr. Lefkofsky makes aspirational claims to investors by tossing in comparisons such as Nvidia and Tesla. He suggests that Tempus is likely to reach a similar inflection point, and that vast upside is around the corner. However, ten years since being founded in 2015, there is no evidence that Tempus has generated a profit or net positive cash flow. In contrast, after a decade, both Nvidia and Tesla had already reached $2 billion in revenue (more than 2x TEM’s revenue) and both were able to have at least one cash flow positive year.

  • Our concerns about the Company’s AI capabilities deepened after we evaluated split-second product demonstrations from promotional videos and screenshots from the Company’s website. Apparently, Tempus has not properly figured out how to leverage AI to cross reference and check even basic facts such as patient age and the sequencing order of samples and deliveries.
    

• Multiple Board members and other executives have been associated with troubled companies that restated financial results.

  • Tempus’ Chief Financial Officer worked in a variety of accounting roles at Groupon, which, as we previously indicated, had challenges. The Chief Accounting Officer (“CAO”) was previously CAO at RTI Medical (renamed: Surgalign) which had a multi-period financial restatement and resulted in revenue recognition fraud allegations by the SEC, though he was not specifically named in the compliant and some of the misconduct pre-dated his arrival.

  • Tempus’ Audit Chairman is the former CEO and Chairman of InnerWorkings, a company founded by Eric Lefkofsky, which restated results and disclosed a material weakness of controls shortly after his announced departure. Also on the audit committee is Peter Barris. Mr. Barris and his investment firm NEA were involved with Vonage Holdings and Neutral Tandem which rebranded as Inteliquent. Vonage, which was led by an SEC-charged executive named Jeffrey Citron, later restated its financials in 2008 while reporting a material weakness of internal controls. Inteliquent commenced an internal investigation of its financial reporting and accounting practices and reported a material weakness, but no restatement was made.

  • Lastly, board member Ted Leonsis has partnered with Mr. Lefkofsky before at Groupon. His biography in Tempus’ SEC filings fails to disclose he was Vice Chairman of AOL during a controversial period at the company and through its merger with Time Warner. Mr. Leonsis made a public statement that he was “very comfortable” with the company’s accounting practices. The SEC later charged eight executives with accounting fraud involving round tripping of revenue and imposed a $300 million civil penalty. Mr. Leonsis was not one of the executives charged.
    

• We see signs of aggressive accounting and financial reporting.

  • The Company’s revenue quality should be carefully scrutinized. To illustrate this, there is evidence of a well-timed announcement to support the Tempus equity story created by aggressive financial engineering and a suspicious joint venture (“JV”) between Japan’s SoftBank and Tempus (“SB Tempus”). The JV was announced on June 27, 2024, shortly after the IPO on June 14, 2024. A SoftBank affiliate invested $200 million in Tempus pre-IPO Series G-5 convertible preferred stock on April 30, 2024. Under the terms of the JV agreement, both Tempus and SoftBank each contributed ¥15.0 billion (approximately $95 million) to the joint venture. Tempus then recognizes revenue and other income from the JV over a 2- to 3-year period. This maneuver has the appearance of potentially round-tripping capital to create revenue and income for Tempus and undermines the credibility and substance of the arrangement. Curiously, the JV’s recently appointed CEO departed within six months after the formation of the partnership.

  • We believe the accounting treatment of the $250 million Google Cloud debt raises significant concerns about the quality of its Adjusted EBITDA and the Company’s path to meeting its FY25 guidance. Specifically, Tempus nets the debt’s principal modification against cloud and software expenses in SG&A based on usage. This non-cash operating benefit, totaling $25.6 million over the last twelve months, directly inflates reported Adj. EBITDA. Without this adjustment, LTM Adj. EBITDA would have been ($102.6) million, rather than the reported ($77.0) million, and Adj. EBITDA margin would be -320 bps lower. We do not know the exact usage mechanism driving the debt principal reduction, but Tempus could be creating a ‘cookie jar’ to achieve its inflection to $5 million of positive Adj. EBITDA for FY25. We believe the non-cash gain should be recorded as “other income” instead of as an operating income benefit.

  • We believe the Total Contract Value (“TCV”) that Tempus reports is aggressively defined with non-binding opt-ins, improbable milestone payments, related-party transactions, and self-funded commitments. For example, $300 million in cumulative opt-ins and $22.4 million in unlikely milestone payments are included in TCV, despite not representing firm contractual commitments. The April 2025 Pathos deal, which accounts for $200 million of TCV, involves a company founded, managed, and funded by Tempus leadership and Board-related investment firms, raising related-party and conflict-of-interest concerns. We also believe the SoftBank JV allowed Tempus to recognize $95 million in TCV which is precisely the amount of capital Tempus contributed.
    

• Key strategic partnerships and deal announcement with AstraZeneca and Pathos AI merit scrutiny.

  • Tempus’s partnership with AstraZeneca began in November 2021 with the announcement of a five-year, $200 million Master Services Agreement (“MSA”), which was positioned as a cornerstone relationship for the Company. Over time, this agreement was amended to extend the term to nearly seven years and increase the total commitment to $220 million. However, beneath the surface, we believe the partnership has weakened and is based on suspicious economic considerations: 1) AstraZeneca ultimately declined to increase its commitment before the end of 2024, forfeiting a warrant to purchase $100 million of TEM stock at the IPO price one year before expiry. We believe this refusal signaled AstraZeneca’s reluctance to deepen its financial stake and pointed to a deterioration in the strength of the relationship, despite public messaging that suggests otherwise, and 2) a provision links AstraZeneca’s increased potential commitment to a doubling of TEM’s share price and not to actual demand for, or perceived value from, the products and services delivered by Tempus. This provision does not, in our opinion, hold up under scrutiny.

  • Pathos AI (“Pathos”) is a Lefkofsky-related company being touted as a unicorn with a $1.6 billion valuation. Ironically, Tempus already owns warrants to purchase 15% of Pathos, but internally appears to ascribe it as worthless on the balance sheet and appointed a Chief Scientist who was the founding CEO of Sema4 (now GeneDx). According to a recent report by Grizzly Research entitled “Insiders Attest That GeneDx Is Actively Committing Widespread Fraud”, CEO Schadt was aware that the company was billing insurance providers for tests without sufficient evidence to support the charges. In late April 2025, Tempus promoted “Expanded Strategic Agreements with AstraZeneca and Pathos” which we believe embodies the extreme financial engineering now characterizing recent deals. Public statements could lead investors to believe this was an incremental, expanded contract from AstraZeneca. We believe the $35 million payment from AstraZeneca to Tempus is not new revenue, but rather a pass-through payment that Tempus is contractually obligated to forward to Pathos as a "Statement of Work" under the original 2021 MSA. In other words, it was already part of AstraZeneca’s existing contractual commitment. By routing these funds to Pathos, we believe AstraZeneca reduced its remaining obligation to Tempus by $35 million and thus was not an expansion of its partnership.

  • The $200 million data licensing and model development deal raises profound conflict-of-interest concerns and calls into question the independence and economic substance of this headline contract. Up to half of the $200 million may be paid in illiquid Pathos Series D Preferred Stock - not cash - and Tempus is required to pay Pathos $35 million as part of the same arrangement. This convoluted structure, involving related parties and non-cash consideration, raises serious questions about the authenticity of the Company’s reported revenue and backlog. Far from being a true extension of the AstraZeneca relationship, we believe the Pathos deal underscores a weaker commitment from a significant customer, while highlighting a reliance on related-party transactions and creative deal structuring.
    

• We believe the Company’s recent financial guidance revision reveals weakness in core operations.

  • We believe that the $10 million FY25 revenue guidance increase earlier this month is substantially explained by the expected contributions from the Ambry Genetics acquisition and the new Pathos/AstraZeneca deal, two factors that we estimate should have added nearly $46 million to projected revenue for the year. With overall guidance rising by only a fraction of this amount, we believe it reveals a significant shortfall in the core genomics and data businesses. This discrepancy strongly suggests that the Company’s fundamental operations are underperforming relative to prior expectations. It further supports the view that headline growth is being driven by acquisitions, related-party transactions, and financial maneuvers, rather than by robust organic growth in its main business lines.

  • There are other indicators of a deterioration in business such as an exhaustion of nearly all long-term deferred revenue and a decline in the growth rate of cloud storage costs as well as a recent amendment to the Google contract.
    

• We believe that owning shares of Tempus is a poor risk / reward.

  • We believe the Tempus equity growth story is built on hype and an appeal to retail investors that it is an exciting and disruptive technology play with AI appeal which could have the next Tesla- or Nvidia-type inflection. However, we think investors should focus on its aggressive accounting, financial engineering, related party dealings, and earnings quality.

  • Tempus is extremely dependent on keeping the perceived value of its equity high because it is among the most aggressive issuers among its peers of stock-based compensation at 77% of revenue. Key insiders and early institutional investors have quickly begun to liquidate stock while its shares command a premium multiple based on its above-industry average revenue growth irrespective of its source and quality. Debt is now rising from the Ambry Genetics acquisition, and we find an unusual financial covenant for a $12 billon public company which requires the Company to maintain a minimum of $1 billion in revenue which we believe puts greater pressure on the Company to be creative in seeking new business.

  • Given these risks, we believe that investors should exercise extreme caution and scrutinize the Company’s public statements closely. The sell-side consensus price target is $66.82 per share which implies just 1% upside potential. We value the core, related party, and acquired revenues of Tempus separately and based on our 2025E estimates, we see 50% – 60% potential downside risk to TEM’s share price. As a result, we do not see a favorable risk / reward in owning the stock and expect shares to underperform the medical diagnostics industry along with the broader equity market.