Canada

As OpenAI and Anthropic formalize their healthcare and life sciences divisions, they are doing more than just selling software; they are auditioning to become the foundational “operating system” for a multitrillion-dollar industry. While their goals are identical — capturing clinical workflows and research cycles — their strategies reveal a fundamental divergence over how a risk-averse sector will ultimately adopt generative AI.

OpenAI: Healthcare at Scale

OpenAI is approaching healthcare as an extension of its broader platform strategy. Its recently announced OpenAI for Healthcare initiative is framed as an enterprise AI stack designed to slot into existing health system workflows, helping organizations automate documentation, reduce administrative burden, and standardize care delivery while meeting HIPAA requirements.

Rather than offering a single healthcare product, OpenAI is emphasizing APIs, business associate agreements, and integrations that allow hospitals, insurers and software vendors to embed its models into clinical decision support tools, chart summarization, care coordination and analytics. The company has highlighted physician-led benchmarking efforts such as HealthBench to demonstrate that its models can meet clinical expectations for reliability and alignment.

At the same time, OpenAI is leveraging a consumer channel that few competitors can match. ChatGPT is already used at massive scale for health-related questions, from interpreting lab results to understanding symptoms and insurance options. That usage is now being formalized through ChatGPT Health, which allows users to securely connect personal health data so responses can be grounded in individual context.

On Monday (Jan. 12) OpenAI announced in a post on X that it acquired healthcare startup Torch, which unifies lab results, medications and visit recordings, and will combine it with ChatGPT Health. In its own blog post about the acquisition, Torch that by bringing together health information that is otherwise scattered, it builds “a medical memory for AI” that helps patients see the whole picture. OpenAI said in its post: “Bringing this together with ChatGPT Health opens up a new way to understand and manage your health.”

OpenAI is careful to position the consumer experience delivered by ChatGPT Health as informational rather than diagnostic, and it is not regulated under HIPAA. Still, the company has acknowledged that healthcare is already one of ChatGPT’s largest use cases, with tens of millions of health-related queries flowing through the system daily, as reported by PYMNTS. That demand effectively serves as a top-of-funnel for enterprise adoption, familiarizing patients and clinicians alike with AI as a default interface for medical information.

Anthropic: Bet on Specialization

Anthropic is taking a more targeted approach. Instead of extending a mass-market assistant into healthcare, it is building healthcare and life sciences offerings on top of the Claude model family, with an emphasis on tightly controlled, domain-specific deployments.

The company’s Claude for Healthcare product is designed for clinicians, insurers and healthcare administrators, with HIPAA-ready infrastructure and direct integrations into authoritative datasets such as ICD-10 coding systems, CMS coverage data, the National Provider Identifier Registry, and PubMed. These connectors enable concrete workflows like prior authorization, report generation, and medical coding interpretation, rather than broad conversational use.

Anthropic is also pushing deeper into life sciences. Through Claude for Life Sciences, the company is positioning its models as research partners embedded in scientific environments, connected to platforms like PubMed, Benchling and ClinicalTrials.gov. The focus is on tasks such as literature synthesis, hypothesis generation, clinical trial planning, and regulatory documentation, placing Claude closer to the core of biomedical research rather than at the patient-facing edge.

A central theme in Anthropic’s messaging is control. The company emphasizes that data accessed through healthcare and research integrations is not used to train its models, and it highlights customizable agent skills, including FHIR-based tool building, that allow organizations to define how Claude operates within strict boundaries. This reflects a view of healthcare as a market where trust, auditability and predictability matter more than rapid experimentation.

For all PYMNTS AI coverage, subscribe to the daily AI Newsletter.

That adaptability is a central theme of “Financial Scams and Consumer Trust,” a November report produced by PYMNTS Intelligence in collaboration with Block.

Based on a U.S. census-balanced survey of more than 15,000 consumers, the research finds that nearly one in five U.S. adults has experienced at least one scam in the past five years, underscoring how deeply fraud has embedded itself in everyday financial life.

Younger, Digital-First Consumers Are Most at Risk

One of the report’s most striking findings upends long-held assumptions about who is most vulnerable. Younger generations, not older ones, face the highest exposure. About 24% of millennials and 22% of Generation Z consumers report having been scammed in the past five years, compared with 14% of baby boomers and seniors.

College-educated consumers are also more likely to be victims than those without a degree, reflecting greater digital engagement and exposure.

The channels matter. Email and phone calls remain the most common entry points overall, but social media plays an outsized role for Gen Z, accounting for nearly one-quarter of first scam contacts among that cohort. Across age groups, fraudsters overwhelmingly rely on impersonation, posing as trusted companies, banks, government agencies or even personal contacts.

Technology as an Enabler for Scammers

Technology has lowered both the cost and the speed of fraud. Scammers increasingly use digital marketplaces, peer-to-peer payment rails and AI-enabled impersonation tools to make schemes more convincing. Speed is a core weapon: nearly two-thirds of victims make a payment within 24 hours of first contact, and many do so within minutes. In more than half of cases, victims send money directly; in the rest, they unknowingly provide account credentials that allow funds to be drained.

The Top Five Scams Targeting Consumers

• Fake debt collection (18%): Fraudsters pose as collectors demanding immediate payment. Victims span income and age groups, with strong impact across generations.
• Scams leading to identity theft (16%): Stolen personal data is used to access or open accounts. These scams are especially damaging and frequently reported to banks.
• Gift card scams (15%): Victims are pressured to buy and share gift card codes. Younger consumers and bridge millennials are disproportionately affected.
• Fake eCommerce or marketplace scams (14%): Nonexistent goods are sold online, with older consumers more likely to report these as their most costly incidents.
• Investment scams (8%): Though less common, they cause the largest losses, with median household losses exceeding $3,000 and heavy use of cryptocurrency payments.

How Technology Can Also Be the Defense

The report makes clear that prevention and response are inseparable. Consumers who report scams to their financial institutions are far more likely to recover funds, and trust rebounds sharply when recovery occurs.

For banks and FinTechs, the defenses increasingly hinge on technology: real-time transaction monitoring, confirmation prompts that slow high-risk payments, stronger identity verification and clearer reporting pathways inside apps and online banking portals. Just as important is education that helps consumers recognize impersonation tactics and urgency cues before money moves.

Fraud may be relentless, but the data shows it is not unbeatable. Faster detection, clearer communication and earlier intervention can reduce losses, preserve trust and keep consumers engaged in the digital economy.

At PYMNTS Intelligence, we work with businesses to uncover insights that fuel intelligent, data-driven discussions on changing customer expectations, a more connected economy and the strategic shifts necessary to achieve outcomes. With rigorous research methodologies and unwavering commitment to objective quality, we offer trusted data to grow your business. As our partner, you’ll have access to our diverse team of PhDs, researchers, data analysts, number crunchers, subject matter veterans and editorial experts.

Canada’s payments landscape is being recast at high speed, and the once-predictable routines of corporate treasurers are being pushed into the digital fast lane.

The change is as much cultural as technological for the practitioners who sit at the center of cash management.

“I have been on the ground here in Canada … for over 25 years,” Lyndsay Langford, Bank of America’s head of Global Payment Solutions for Canada, told PYMNTS. “In the past, we had wires, EFT and checks. Today, clients demand faster payments, less friction and richer remittance data.”

Those demands have forced treasury teams to abandon manual processes and build real-time visibility into positions and forecasts. The pandemic “shifted us toward a more digital mindset,” accelerating projects that once sat on multiyear road maps, she said.

That shift is playing out against a sweeping upgrade of Canada’s underlying rails. Payments Canada began its modernization plan in 2016, replacing the legacy wire system with Lynx and preparing the country’s first real-time rail (RTR).

“We’re heavily focused right now as a country on the buildout and the launch of real-time payments,” Langford said.

Once live, RTR will settle payments in seconds, carry data-rich messages and eventually enable QR-code requests for payment and cross-border links to other real-time payment methods. Interac — once a person-to-person tool — now supports near-instant business-to-consumer disbursements. Two years ago, Bank of America folded the channel into its global digital disbursements platform alongside Zelle and PayPal.

The glue that makes many of those ambitions useful to a treasurer is ISO 20022, a global messaging standard that expands the amount of information that can travel with each transaction. Bank of America has been running a formal ISO migration program since 2019 and has already completed “over 10 clearing migrations around the world,” Langford said.

Early Canadian adopters are exploiting the richer data to “remove a lot of manual work from the reconciliation process,” she said because ERP systems can auto-match invoices to payments the moment funds hit. The bank is building direct ISO feeds into corporate accounting platforms to maximize straight-through processing.

Risk Mitigation

Yet data alone does not erase the uncertainty of sending money across borders. Pandemic shocks, geopolitical tension and exchange rate swings have only amplified that uncertainty.

“Risk mitigation for cross-border payments is rising to the top of the priority list,” Langford said.

Bank of America’s answer includes a guaranteed foreign exchange (FX) rate solution that allows clients to lock in an FX rate for up to a year — “market leading” in tenor, she said — and CashPro Forecasting, a machine learning tool that predicts liquidity needs so companies can “make more intelligent working capital decisions” and pivot when volatility strikes.

Speed, meanwhile, is not always paramount.

“We want to make sure the options we put in front of clients are cost‑effective and fast where speed is important,” Langford said.

Still, the bank also aims for a globally consistent user experience so a Canadian treasurer responsible for Asia or Europe sees the same interface and controls.

Much of that experience is delivered through CashPro, Bank of America’s digital treasury suite used by more than 40,000 corporate clients worldwide. When the bank combined its Global Transaction Services and Enterprise Payments teams under the banner of Global Payment Solutions in 2023, it created a pipeline for ideas developed in the retail bank to flow into the commercial platform. The most visible import is CashPro Chat, built on the artificial intelligence backbone that powers Erica, Bank of America’s consumer virtual assistant. The tool lets treasurers ask questions and receive 24/7 answers from an “intelligent virtual service advisor,” escalating more complex issues to human specialists, Langford said.

“The more we collectively use it, the more it learns and grows,” she said.

Integration has also sped product rollouts. A U.S. electronic payments collections service that lets businesses accept card and ACH payments will soon debut in Canada after clients “raved about it for years,” she said, illustrating how the new structure “allows us to consider how investments can be leveraged more broadly.”

Listening to the Pros

Listening to clients is as important as engineering. CashPro advisory boards meet regularly “with clients around the world, and that includes right here on the ground in Canada,” ensuring new features target the highest priority wishes, Langford said.

The next inflection point is already on the horizon. The Retail Payment Activities Act will allow nonbank providers to participate directly in the RTR, intensifying competition and giving corporates more ways to customize how they pay and get paid.

“With this complexity, … the goal is to make things easier as they become more complex,” Langford said.

She said she expects that competitive pressure — and the data fabric of ISO 20022 — will trigger a wave of innovation in financing, supply‑chain optimization and value-added analytics.

“The real-time rail will be the platform for future innovation in Canada,” Langford said. “It’s going to increase competition, and when you have increased competition that benefits all Canadians — consumers and commercial entities alike.”

For treasurers, the tools are arriving quickly, and those who master them will turn liquidity management from a back-office chore into a source of strategic advantage.

“We need to be at the table — and we are at the table — to ensure our clients’ voices drive the next three- to five-year journey,” Langford said.

The pay later company and the travel search engine have offered services to travelers in the United States since 2023 and are now bringing their collaboration to Canada, Kayak said in a Thursday (May 22) press release.

“Consumers are increasingly turning to Affirm when booking their flights, hotels, rides and more as flexible payment options remain a top priority for travelers across Canada,” Affirm Chief Revenue Officer Wayne Pommen said in the release. “This expansion with Kayak is a natural next step for our long-standing partnership as we look to offer even more travelers peace of mind when paying for their next trip using Affirm.”

By choosing Affirm at checkout on Kayak’s Canadian website, approved travelers can split the cost of flights, accommodations, and car rentals or car sharing into monthly payments, according to the release.

Affirm already works with several other travel companies, including Booking Holdings brands Agoda, Booking.com and Priceline, which itself owns Kayak.

Kayak Chief Product Officer Matthias Keller discussed with PYMNTS Thursday the company’s launch of Kayak.ai, a conversational travel booking assistant powered by generative artificial intelligence.

“This is not just a chatbot,” Keller said. “It’s a ChatGPT just built for travel. You can ask any travel question, and in a conversation, you get real-life rates that are also bookable on Kayak.”

Kayak.ai was created to emulate natural human conversation and the kind of planning a traveler might do with a knowledgeable agent or friend. It isn’t fully automated, but future versions of the assistant will allow for seamless and even semi-autonomous bookings within the chat itself, Keller said.

“Right now, it’s more like an attended booking within the chat,” he said. “You’ll still see what you’re booking, confirm the price, and maybe enter your credit card. But over time, this could get more and more semi-automated. Maybe you leave your details, and the booking happens in the background.”

Meanwhile, Kayak’s integrations with major airlines, hotel chains and global distribution systems (GDS) offer it a competitive edge, Keller said.