Windows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallCrashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minuteiTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
Yes, according to SAP’s chief AI strategy officer Sean Kask, but as a strategic warning rather than a measured forecast. Speaking at Wave by Vento in Turin in a CNBC session moderated by Carolin Roth, Kask said: “Like the famous quip that every company is becoming a software company, every software company has to become an AI company, or they’ll perish.” The remarks were reported by The Next Web on October 8, 2026. The article gives no data showing how many software firms are at risk, so the “perish” language should be read as rhetoric about competitive pressure.
What Kask actually argued
Kask’s core point is that established software companies should rebuild their products around AI instead of adding AI as one more feature. The reporter paraphrases this as companies needing to “rebuild their products around AI, not add it as one more feature.” That paraphrase is not a direct quotation, so the sentence quoted above is the one to cite.
The argument is about product design and workflow. A chatbot bolted onto an existing menu is one thing; changing how a user gets from a question to a result is another. Kask’s argument is about the second.
Do these 3 things before closing this tab:
1Repair Windows errors before they cause bigger problems2Fix the driver behind crashes, sound loss and screen glitches3Clear out junk files and repair common Windows errorsWhat an AI-first product looks like
The example Kask gave is an interface where a user asks a question in plain language and the system generates the tables they need on screen. The interaction replaces a sequence of menus, filters and report builders with a single request.
#1 Best Overall
That example is the only concrete illustration reported. It supports a reading of “AI-first” as a change in how the product works, but SAP has not published a formal definition of the term in the coverage reviewed. Kask also tied this kind of interface to structured enterprise data: the system is only useful when the underlying tables are organized and linked.
Why SAP is betting on tables
Kask separated two kinds of model. Language models learn by predicting text. Tabular models handle structured rows and columns and are used for numerical prediction and classification. He argued that tables carry a disproportionate share of business value and that language models alone do not serve them well.
SAP-RPT-1 and internal use
According to the report, SAP has built tabular models internally for a couple of years and Kask said SAP uses SAP-RPT-1 in production. The article does not describe deployment scale, customers or measured results for that model.
The Prior Labs acquisition
The report says SAP’s reasons for acquiring Prior Labs were its tabular-model work and overlapping research with SAP. Kask reportedly said Prior Labs’ model led the TabArena benchmark and had applications in cancer diagnosis and bank transactions. Those are reported claims. The article does not include the benchmark results, the methodology, or any transaction documents.
The deal is reported as completed in July 2026, with SAP committing “more than €1 billion over four years” to develop Prior Labs into a frontier AI lab. According to Kask, SAP will keep Prior Labs as a separate lab with room for its own research, and will keep its model open weight so researchers and startups can use it. The €1 billion figure is attributed to The Next Web’s reporting; the article includes no direct quote or deal documentation for it.
Foundation models versus task-specific models
Kask’s most practical claim is that a single foundation model can do in days work that used to take teams weeks or months of training and tuning many task-specific models. He also said it beat methods such as XGBoost on accuracy. The report gives no test setup, dataset or scope for that comparison, so treat it as a claim about SAP’s experience rather than a benchmark result.
Rank #3
Why SAP does not build its own large language model
According to the article, SAP does not build its own large language model, citing cost and convergence. Kask said customers can use models from Google, OpenAI, Anthropic and Mistral. SAP uses more than 100 models internally and tests each use case against them to choose the best fit.
The Tool Desk
Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Outbyte PC Repair FREERepair Windows errors before they cause bigger problemsFix Now →His wider position is that Europe should avoid competing mainly in the crowded race to build ever-larger language models. He called that race a “red ocean” and argued that the more useful assets are enterprise data and systems. As evidence he cited SAP’s knowledge graph, which he said links 500,000 tables and 7 million fields and helps make agents accurate enough for workplace use. The report also notes that Mistral had launched Large 4, described there as a one-trillion-parameter open-weight model. The report does not establish that model’s benchmark standing.
The figures Kask used, and how to read them
Kask’s numbers were stated at the event and are reported by The Next Web. The article does not explain how they were calculated, so cite them as statements made by SAP’s executive rather than as sector-wide statistics.
- “About 80% of business data is unstructured.” Kask, 2026.
- “The 20% held in tables generates 80% of a company’s value.” Kask, 2026. The report gives no source or method for the value split.
- Knowledge graph scale of “500,000 tables and 7 million fields”. Kask’s description of SAP’s own graph, not an independent count.
- “Every dollar of SAP software sold generates $6 to $10 for its ecosystem of partners.” Kask, 2026. This is a claim about partner economics that the article does not substantiate.
- “More than 100 models internally.” Kask’s description of SAP’s internal model use, with no list or dates given.
Strategic options compared
The article describes two decisions every enterprise software vendor faces: whether to build a general-purpose language model or choose among providers, and whether to add AI to an existing workflow or redesign the workflow around it. The table below sets out the position Kask described and what the report does not establish.
| Decision | Kask’s stated position | What the report does not establish |
|---|---|---|
| Build a general-purpose language model | Not SAP’s approach, citing cost and convergence | Cost figures, convergence evidence, or any SAP model roadmap |
| Select among third-party models | SAP tests more than 100 models per use case; customers can use Google, OpenAI, Anthropic and Mistral | Evaluation protocol, selection criteria, or provider roster dates |
| Add AI as a feature | Described as insufficient for established software firms | Adoption rates, revenue effects, or failure rates for firms that take this path |
| Redesign the workflow around AI | Plain-language requests that generate tables on screen | User outcome measures or adoption data for that interface |
| Use tabular models for structured data | Tables carry disproportionate value; SAP-RPT-1 in production; Prior Labs acquired | Benchmark methodology, deployment scale, or customer results |
Questions to ask before following the strategy
Kask’s argument is a direction, not a checklist, but a software team can test it against its own products. These questions separate a real workflow redesign from a feature rename:
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
- Does the user reach a result in fewer steps than before, and can you measure that difference?
- Is the target task supported by structured data your systems already hold, such as tables with consistent keys and definitions?
- Have you compared several models on that specific task, rather than picking one because it is widely known?
- Can you keep control of the data and the model choice if a provider changes terms or pricing?
- Are the accuracy claims backed by an evaluation you can reproduce, not only by a vendor statement?
The Next Web’s full report, including Kask’s remarks in context, is at thenextweb.com/news/sap-sean-kask-prior-labs-every-software-company-ai-company-wave-by-vento. SAP has not been verified as the source of the figures in that report through an official release in the coverage reviewed, so check SAP’s own announcements before citing the numbers as company disclosures.
Reported by Ana Maria Constantin, The Next Web, October 8, 2026.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
