Blog

Why Thomson Reuters Risk & Fraud Solutions?

After a comprehensive process evaluating alternative solutions, Palmar Forensics selected Thomson Reuters CLEAR as the backbone of its investigative and vendor credentialing workflows for several reasons.

The first was comprehensiveness and navigability; full reports are thorough yet easier to navigate than alternatives. Their prior tool generated 200 to 300-page reports with lots of noise; CLEAR’s reports contain less irrelevant information, are structured, and searchable. As Joseph Palmar, CEO of Palmar Forensics puts it, “CLEAR is more effective and efficient. The CLEAR report is still 60 or 70 pages long, but it’s simple to navigate. It’s just so much easier to find out what you need to know. And it’s complete and accurate — it’s so comprehensive.”

CLEAR’s algorithms and logic were also praised. Particularly the strong relationship and ownership discovery that uncovers shell or sister companies and bid-rigging; targeted modules like CLEAR Risk Inform, CLEAR ID Confirm, and seamless alerts for continuous monitoring. These features gave CLEAR a competitive advantage.

Lastly, Palmar called out the credibility of Thomson Reuters and CLEAR within the industry, making it easy for him to trust.

“The Thomson Reuters CLEAR solution is the gold standard of fraud investigations. Law enforcement and the federal government both use it. I’ve lost count of how many meetings I’ve walked into where we’ve mentioned that we use CLEAR and they instantly realize we know what we’re talking about,” he says.

The value of CLEAR

Since 2012, Palmar Forensics has been using CLEAR — which provides comprehensive, up-to-date information from public records, sanctions databases, and media coverage. It also performs ID validation through CLEAR ID Confirm and subject-risk analytics through CLEAR Risk Inform.

When onboarding new vendors, Palmar Forensics uses CLEAR as the backbone of its proprietary vendor credentialing tool, VETTED®, to verify that the suppliers its clients want to work with are legitimate and independent businesses — and that their officers and or directors are “clean.” CLEAR does so by researching areas such as current and previous addresses, business relationships, criminal records, and creditworthiness. Then, they can be onboarded with greater certainty.

“We make sure clients are dealing with reputable organizations that aren’t shell companies, that aren’t linked to other similar businesses for bid-rigging purposes, and that the company’s officers aren’t criminals or have a nefarious history. In our world, that’s not an unusual situation — it’s an everyday occurrence,” says Palmar.

Uncovering hidden insights

If clients have suspicions about an existing vendor, Palmar Forensics uses CLEAR as part of an automated tool called VETTED to do a deeper dive to check for undisclosed relationships with other organizations, criminal links or backgrounds, or any lawsuits, liens or judgements. In one recent example, an investigation into a list of vendors provided by a client uncovered a conflict of interest in the awarding of contracts — several “separate” vendors were owned by the same people. In another example, one company owner was unaware that their business partner of 20 years had been charged with assault with a deadly weapon.

“Using Thomson Reuters Risk & Fraud Solutions as the backbone of VETTED gives us insights we wouldn’t otherwise have, so clients can understand who they’re doing business with,” explains Palmar. “CLEAR allows us to identify whether there’s anything in a company or individual’s background that’s fraudulent or in any way problematic. We have a suite of tools that we use, but the backbone of everything is CLEAR from Thomson Reuters.”

CLEAR alerts are incredibly useful, according to Palmar. By setting up alerts on certain entities and officers, it is notified about any potential reputational or financial hit to its clients, after the initial onboarding verification process. For Palmar Forensics, this continuous monitoring adds another important layer to the solution’s capabilities.

When considering the value of CLEAR, Palmar says decision-makers need to think beyond strict return on investment (ROI) metrics. That’s because vendor fraud does not usually happen at volume. But when it does occur, the damage can be incalculable. So, while only a small fraction of vendors commit fraud or have criminal associations, organizations must do everything they can to root them out. CLEAR enables them to do that — and that’s where its true value lies.

He cites the example of a charitable foundation where seven of its 500 vendors — accounting for 1% of its vendor base — stole $3 million over the course of five years. In addition to the direct financial loss, this raised the risk that donors would no longer donate money to that organization. Similarly, in the health care sector, if fraud is uncovered in the Medicaid or Medicare space, companies may find their government funding is suspended or withheld, which could have significant implications for business continuity beyond the loss itself.

“You need to spend money to prevent fraudulent expenditures and protect your reputation,” Palmar says. “If you have a massive fraud, your stock price may be affected but there may be even bigger problems to deal with. You can’t measure reputational harm, but one mistake can destroy the reputation you’ve worked so hard to develop.”

“To my mind, it’s better to do the preemptive work to protect the organization. At least if you’ve done your checks, that’s something you can say to the media, or to the board, as opposed to saying no, we didn’t do that because there was no ROI for doing so.”

Unlock the Editor’s Digest for free

Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.

The writer is senior vice-president for research, labs, technology and society at Google-Alphabet

Quantum computing able to harness the subatomic world may seem like a distant prospect to some people. This misapprehension persists despite recent research breakthroughs, many built on the foundational research of 2025 Nobel laureates John Clarke, Michel Devoret, and John Martinis. But quantum computers are already being used for scientific research and real-world applications appear likely to emerge in the near future, even before we get to the goal of large-scale, error-corrected computing systems.

Estimates of how long it will be before we can use quantum computing in ways that will affect most people’s lives vary from a few years to decades. Rather than a single light-switch moment when we shift all at once from not having a quantum computer to having one, it’s more likely that we will follow a more gradual quantum continuum as the systems increasingly become useful.

A helpful way to think about this stage of quantum development is something akin to artificial intelligence in 2010, when neural networks were on the cusp of becoming useful but before advances like protein structure predictor AlphaFold or generative AI chatbots.

Quantum technology holds the promise of solving difficult problems, helping us to learn the structure of quantum systems, from molecules to magnets to black holes.

Similar to early neural networks, quantum computing is already proving useful as a platform for research and experimentation. For example, researchers such as Misha Lukin at Harvard have used them to uncover new phenomena like many-body quantum scars, which reveal order embedded within quantum chaos.

We are also starting to see the potential for real-world applications with quantum advantage (ie problems that quantum computing can solve that are beyond the reach of classical supercomputers), which can advance areas like drug discovery and material science.

We are making progress on quantum computing itself: Google’s quantum processor ran in minutes a benchmark algorithm that would take today’s fastest supercomputers 10 septillion years (a period of time that vastly exceeds the age of the universe). More significant was the demonstration of “below threshold” error correction — a long-standing challenge in the field. Last week, we published the first algorithm with verifiable quantum advantage on hardware and in proof-of-principle experiments with the University of California, Berkeley, found it could be used to understand molecular structure when combined with a common chemistry technique.

Researchers and mathematicians have highlighted at least 70 algorithms — from mathematics and data science to simulation and machine learning — that could allow quantum computers to solve problems much faster than classical ones in a wide range of useful areas. One that has captured public attention since its discovery in the 1990s is Shor’s algorithm, which is expected to be capable of cracking much of the encryption used by today’s most advanced communications systems and digital networks.

While not imminent, here too we are closer than public perception might acknowledge. This year there was a 20-fold decrease in the estimated size of the quantum computer needed to run Shor’s algorithm and crack the widely used RSA-2048 encryption. This shortened timeline underscores the importance of migrating to new, post-quantum cryptography standards, such as those released by the US National Institute of Standards and Technology last year. 

Within five years we are likely to start to see real-world applications become possible, even before we get to a large-scale error corrected quantum computer. In order to be prepared to capitalise on these developments, different sectors have the opportunity to lay the groundwork now.

For governments, that might look like investments in quantum infrastructure such as cryogenic cooling systems, specialised electronics and quantum-ready networks. For businesses, it may be determining how the technology might be useful to their respective industries, for example accurate simulation of molecules in the creation of batteries for electric carmakers, or fusion power for the energy industry.

Individuals should consider how quantum might influence their own careers. We’ll need more specialised engineers (including electrical, mechanical, systems and software engineers), technicians, and applications developers. For anyone thinking about the jobs of the future, these are the fields that could be areas of potential opportunity.

Unlock the Editor’s Digest for free

Roula Khalaf, Editor of the FT, selects her favourite stories in this weekly newsletter.

Europe’s investment banks were already fighting an uphill battle to catch up with their Wall Street rivals, which dwarf them in both size and valuation terms. The return of Donald Trump to the White House has created a medley of additional challenges, laid bare in quarterly earnings this week.

The first problem: deals. In the US, Trump’s deregulatory agenda has encouraged a flurry of mergers and new listings at home. Yet his unpredictable foreign policy is eliciting caution abroad. So while takeovers of US companies jumped 25 per cent in the first nine months of this year to $1.4tn, according to LSEG data, European deals were up just 11 per cent to $554bn. Similarly, new listings in the US have raised more than twice as much as IPOs in Europe, according to Dealogic.

The result is that the high-margin fees that banks mint from corporate activity are harder to come by. Lars Machenil, chief financial officer at BNP Paribas, told analysts that European companies are in “wait and see” mode. 

Despite this, BNP’s US business is growing at a similar rate to Wall Street rivals, Machenil claims. But then comes the second — less intentional — impact of Trump: the dollar. Even after a recent rebound, the currency’s value against major trade partners is down almost 9 per cent year to date. That means European banks’ US earnings are no longer worth as much once translated into their own currency. 

At BNP, foreign exchange moves were a big reason it missed third-quarter revenue forecasts. For Deutsche Bank, currency movements were the difference between expanding its loan book and shrinking it. At UBS, the particularly sharp rally in the Swiss franc earlier in the year unfavourably pumped up the bank’s leverage ratios, making its balance sheet more tricky to manage. 

The other D making some European bankers jealous: deposits. Between geopolitical uncertainty, White House efforts to encourage US investment, and higher interest rates, large companies are moving more of their cash to US accounts. Deposits in Bank of America’s global banking unit, for example, jumped 15 per cent year on year in the third quarter, while JPMorgan’s rose 12 per cent. At Deutsche Bank, in contrast, corporate deposits fell, despite enthusiastic talk of a wave of defence-related investment in the continent.

That’s inconvenient because, as Crisil Coalition Greenwich has highlighted, these sorts of corporate deposits boost the US banks several times over: they earn interest on the deposits themselves, make further money from cash management and payments businesses, and gain sticky funding for other parts of the group. 

An optimistic take on this would be that maybe Europe’s banks aren’t as far behind Wall Street as they look: strip out some currency and other drags and both sides are growing at similar rates. Europeans will benefit proportionally more if the local deal pipeline unclogs. Even then, though Americans will rule the roost. Closing that gap is yet another D: doubtful.

nicholas.megaw@ft.com

Animals

Three-month-old sheep were housed in controlled environmental conditions and divided into two groups. One group of ten sheep were fed control feed (1050 kcal); the other group were fed high-concentrate feed (2010 kcal). The mouse was kept in controlled environmental conditions that were free from specific pathogens and divided into two categories of dietary conditions. One group were fed a standard control diet (4.37 kcal); The other group were fed a high-concentrate diet (16.45 kcal). The animals were euthanized by CO₂ inhalation. During the procedure, CO₂ gas was introduced into the euthanasia chamber at a rate of 10% of the chamber volume per minute (corresponding to a flow rate of 5.8 L/min). The gas flow rate was precisely controlled by adjusting the flow control valve to minimize stress in animals prior to loss of consciousness. CO₂ gas was continuously supplied until the animals completely stopped moving, ceased spontaneous breathing, and exhibited dilated pupils. Once these signs were confirmed, the CO₂ gas supply was turned off, and the animals were observed for an additional 2–3 min to ensure death. All animal experiments were conducted in accordance with the ethical policies and procedures approved by the Inner Mongolia University Animal Care and Use Committee (IMU-MOUSE-2023-013, IACUC Issue No. IMU-SHEEP-2021-011), Inner Mongolia University, China.

Single-cell library and sequencing

Testicular tissues were dissected into small fragments and digested with collagenase IV (1 mg/ml, 37 °C, 15 min). After filtration through a 40 μm cell strainer (BD Falcon, USA, 352340), the resulting single-cell suspension was assessed for viability (>90%) and concentration (>1000 cells/μl). Single-cell libraries were prepared using the Chromium Single Cell 3’ Kit v3 (10 × Genomics, USA) and sequenced (PE150) on an Illumina NovaSeq 6000 platform following the manufacturer’s protocol. The sheep reference genome was processed using cellranger mkgtf and cellranger mkref, and raw data were aligned using cellranger count. Subsequent analysis was performed in Seurat, including data normalization, integration, and differential gene expression analysis (FindAllMarkers, thresholds: |log2FC| ≥ 0.25, P < 0.05). Functional enrichment was conducted using g: Profiler based on Gene Ontology (GO) terms.

Metabolome detection and analysis

Metabolomic profiling was performed on sterilely collected small intestinal contents using a Waters Acquity I-Class PLUS UPLC system coupled with a Xevo G2-XS QTof high-resolution mass spectrometer (Waters). Separation was achieved on an Acquity UPLC HSS T3 column (1.8 μm, 2.1 × 100 mm). Raw data were normalized prior to statistical analysis. Differential metabolites were identified based on significance thresholds of P < 0.05 (Student’s *t*-test), |log₂FC| ≥ 1, and VIP (Variable Importance in Projection) > 1 (derived from multivariate cross-validation). Metabolite enrichment analysis was conducted using MetaboAnalyst 5.0, and visualization was performed in R Studio (v4.2).

16S rDNA amplicon sequencing and analysis

After collecting small intestinal contents under sterile conditions, genomic DNA was extracted and purified using the QIAGEN kit. Subsequently, the samples were sequenced using the Illumina platform.

Subsequent data analysis was conducted utilizing R Studio (v4.2) in conjunction with other relevant software tools. Initially, the data underwent alpha diversity and beta diversity analysis using QIIME2. Subsequently, Metastats (v2020.6) was employed to conduct a t-test on the species abundance data between the groups. Differential species were identified using a stringent dual-filter criterion: unadjusted P value < 0.05 and Q value < 0.05. Species with a p-value less than 0.05 were deemed significantly different. To assess the correlation between genera exhibiting distinct abundances, Spearman correlation analysis was performed, and correlation values with a significance level of P value < 0.05 were retained. Following this, the resulting correlation network was visualized using Cytoscape (v3.9.0). The functional prediction of differential genera is carried out using the Linear discriminant analysis Effect Size (LEfSe) method and the PICRUSt2 tool.

Hematoxylin–eosin (H&E) staining

The testis was placed in Bouin’s solution (Sigma, HT10132, Germany) at 4 °C overnight after the dehydration and embedding of tissue blocks. For this study, serial sections with a thickness of 5 μm were utilized. Following drying at 37 °C, the samples were stained using Lillie-Mayer hematoxylin (Solarbio, G1080, China) and eosin (Zsbio, ZLI-9613, China). The tissue sections were subsequently dehydrated with alcohol and sealed with Neutral Balsam Mounting Medium (ZSGB-BIO, 96949-21-2, China). Finally, the samples were left to dry naturally, facilitating easy observation and statistical analysis.

Immunohistochemistry

The testis was placed in Bouin’s solution (Sigma, HT10132, Germany) at 4 °C overnight after the dehydration and embedding of tissue blocks. The tissue sections were dewaxed and incubated in a blocking solution (3% BSA, 10% normal donkey serum in TBST) for 30 min. The primary antibody (Table S1) was incubated overnight at 4 °C. Subsequently, the corresponding secondary antibody was applied, followed by color development using DAB (ZSGB-BIO, ZLI-9013, China). Finally, counterstaining was performed using Lillie-Mayer hematoxylin (Solarbio, G1080, China), and the samples were dehydrated and sealed for further analysis.

Immunofluorescence

The testis was placed in 4% paraformaldehyde solution at 4 °C overnight after the dehydration and embedding of tissue blocks. The experimental procedures for dewaxing, blocking, and the application of primary and secondary antibodies followed the standard protocols for immunohistochemistry (Immunohistochemistry). Following the completion of the secondary antibody incubation, the nuclei were stained with DAPI. Finally, the samples were sealed using Fluoromount™ Aqueous (Sigma, F4680, Germany) Mounting Medium for further analysis.

RNA extraction and RT-qPCR

Total RNA from testicular tissue and mouse TM4 Sertoli cell line was extracted using the EasyPure RNA kit (TRANS, ER101-01, China). Subsequently, cDNA was synthesized through reverse transcription using an RNA reverse transcription kit. RT-qPCR was performed on a Roche Light Cycler 480II System (Roche, Germany) using TB Green Premix Ex TaqII (Takara, RR820A, China). The primer sequences for this study were provided in Table S2.

Fecal microbiota transplantation

Use 3-week-old male ICR mice as recipients for fecal microbiota transplantation. Prepare a solution by mixing 1 g of sheep small intestinal contents with 1 ml of sterile glycerol (20%) and diluting it with normal saline to a concentration of 0.05 g/ml. Mice are treated with a combination of antibiotics, including vancomycin (0.5 g/L), neomycin sulfate (1 g/L), metronidazole (1 g/L), and ampicillin (1 g/L), for two consecutive weeks prior to microbiota transplantation. After the antibiotic treatment, fecal transplantation experiments were conducted daily for a duration of eight weeks.

Cells culture and treatment

Mice TM4 Sertoli cell line were cultured in DMEM medium containing 10% fetal bovine serum (Gibco, 10099-141, Australia) and 1% penicillin-streptomycin (HyClone, SV30010, China). The cells were cultured in a cell incubator at 37 °C and 5% CO₂, and the medium was changed every 24 h. Cells were passaged when the cell density reached 80–90%, and cells were treated 24 h later (5 mM L-Citrulline (MCE, HY-N0391, China); 35 μM Sucrose; 35 μM Sucrose + 5 mM L-Citrulline).

Cell proliferation assay

Cell proliferation was detected using MTT Detection kit (Beyotime, C0009, Nantong, China). After 48 h of treatment, discard the medium, add 10 μl/well of MTT working solution into a 96-well plate, and incubate at 37 °C for 4 h in the dark. Next, add 100 μl per well of the Formazan solution from the MTT kit and incubate for 1 h in the dark. Finally, the absorbance of the 96-well plates at 570 nm was measured using a Microplate Reader (Bio-Rad, iMark™, USA).

ROS evaluation

For TM4 cell line, ROS staining reagent (Beyotime, S0033S, China) was added to a 96-well plate. Then, the plate was incubated at 37 °C for 30 min in a dark environment. Following this, the cells were washed twice with PBS and the fluorescence intensity of each group was immediately observed using a Nikon fluorescence microscope. For tissue ROS assessment, tissue homogenates were prepared in liquid nitrogen and centrifuged to obtain the supernatant. ROS levels were measured using a mouse ROS ELISA kit (JILID, J25190, China). Following the addition of enzyme conjugate reagent, the plates were incubated at 37 °C for 60 min. After washing with the provided buffer, the wells were blotted dry and developed with substrate at 37 °C for 15 min, protected from light. Absorbance was measured at 450 nm.

Sertoli cell isolation and culture

Testes from 7-day-old mice were decapsulated and digested sequentially. The first digestion used 0.1% collagenase IV at 37 °C with shaking (85 rpm) for 20 min. After sedimentation, the supernatant was discarded, and the tissue was washed twice with PBS. The second digestion was with 0.1% collagenase IV and 0.1% hyaluronidase at 37 °C for 10 min, followed by centrifugation (500 rpm, 1 min) and PBS washing (x3). The third digestion involved 0.1% collagenase IV, 0.1% hyaluronidase, and 0.25% trypsin at 37 °C for 20 min. After terminating the digestion, the solution was centrifuged (1000 rpm, 3 min), and the pellet was resuspended in culture medium. The single-cell suspension was seeded onto Matrigel-coated dishes and cultured at 34 °C with 5% CO₂ for 48 h. Cells were treated with 20 mM Tris for 2.5 min, washed with DMEM/F12 (x3), and culture was resumed. Drug treatments followed the same protocol as for TM4 cells.

Western blot

The TM4 Sertoli cells were lysed using RIPA buffer (Beyotime, P00113B, China). Proteins were separated by SDS-PAGE and transferred onto PVDF membranes (Millipore, ISEQ00010, Germany). After blocking with 5% BSA for 1 h, the membranes were incubated overnight at 4 °C with primary antibodies (Table S1), followed by a 2 h incubation with HRP-conjugated secondary antibodies at room temperature. Protein bands were visualized and quantified using ImageJ software.

Enzyme-linked immunosorbent assay (ELISA)

The concentrations of L-citrulline in the small intestinal contents and testis were quantified using enzyme-linked immunosorbent assay (ELISA) kits (J43384 and J25562, Giled Biotechnology, Wuhan, China), following the manufacturer’s protocols.

Statistical analysis

In this study, GraphPad Prism (v 8.0) software was used for statistical analysis. Bioassay data were shown as mean ± SEM. Data analysis was performed using two-tailed Student’s t test or one-way analysis of variance (ANOVA). Significant difference was *P < 0.05, **P < 0.01, ***P < 0.0001. Two-tailed Student’s t-test was used to compare the mean differences between two independent samples. One-way ANOVA was used to compare the mean differences among four independent samples, with post-hoc analysis performed using the Tukey’s test.