Do these 3 things before closing this tab:
1Repair Windows errors before they cause bigger problems2Scan for outdated or missing drivers - takes under a minute3Clear out junk files and repair common Windows errorsSome companies are looking beyond frontier AI providers because they want tighter control over sensitive data and proprietary information—not because open models are automatically safer or better. A new Fortune report describes a growing range of approaches, from enterprise privacy terms to cloud intermediaries and company-run models. It does not show that businesses are broadly abandoning OpenAI or Anthropic; many are choosing different arrangements for different tasks.
Why data privacy is changing AI vendor decisions
For companies using AI, the concern is not only whether a provider trains on submitted data. Executives also weigh data retention, exposure of intellectual property, and dependence on an outside provider. Fortune’s October 5, 2026 report says OpenAI and Anthropic have said they do not train on enterprise data. Security experts interviewed by Fortune nevertheless raised concerns about broader ways a provider might learn from customer operations. Those concerns are expert opinions, not evidence that either company misused customer data.
As an Amazon Associate I earn from qualifying purchases.
Optiv vice president and chief information security officer Rob Gregory summarized the perceived risk this way: “From a business standpoint, the risk of any of that data being trained on an AI model is unacceptable.” Microsoft CEO Satya Nadella, quoted by Fortune, framed the tradeoff as paying for intelligence with money and potentially with proprietary knowledge that a company must reveal to make the service useful. These remarks describe the stakes companies are weighing, not a finding that sharing data necessarily causes it to be reused.
Recommended Free Tools
What changed in the OpenAI and Anthropic privacy debate?
Fortune reports that Anthropic’s June 2026 announcement of 30-day retention for chats with its Fable and Mythos models prompted controversy. The report also says OpenAI restated an enterprise zero-data-retention offer in an August 19, 2026 blog post and previewed Private Safety Processing, described as an option for customers to store data in their own cloud. Anthropic announced a similar own-cloud option on September 1, 2026, according to Fortune. The story also reports, citing The Information, that Booz Allen restricted employee use of Fable.
#1 Best Overall
- Built for Local AI Development: AMD Ryzen AI Halo is designed for local AI development and inference, featuring 128GB unified memory and support for up to 200B parameter models to build and run intensive AI workloads locally.
- 128GB Unified Memory: Features 128GB LPDDR5x unified memory at 8000 MT/s with 256 GB/s memory bandwidth, providing a shared memory pool across the CPU, GPU, and NPU to support larger AI models.
- AMD Ryzen AI Max+ 395 Processor: Features 16 cores, 32 threads, and Zen 5 architecture, paired with AMD Radeon 8060S integrated graphics featuring 40 RDNA 3.5 compute units and an AMD XDNA 2 NPU with up to 50 TOPS.
- Linux AI Developer Platform: Purpose-built for Linux-based AI development with full AMD ROCm software support and preloaded tools, models, and workflows optimized for local AI development.
- Compact, Connected Design: Includes a 2TB M.2 SSD, 10GbE LAN, Wi-Fi 7, Bluetooth 5.4, USB-C connectivity, and HDMI 2.1b.
These are dated developments as reported by Fortune; the report does not independently establish the terms, eligibility, or practical limits of each offering. Organizations should evaluate the provider’s current contract and product documentation for the specific plan, model, retention settings, and data flows they would use rather than assuming that an enterprise label guarantees a particular arrangement.
What does “sovereign AI” mean?
In Fortune’s account, sovereign AI means having control over more of the AI stack. That can mean using downloadable open models in a company’s own cloud, operating proprietary cloud infrastructure, or owning the chips that run the workloads. The term was historically associated with governments, but the report says it is increasingly used in corporate discussions.
It is a spectrum, not a single product or a guarantee of privacy. A company can retain more control over where data is processed and which model it uses without owning every layer of infrastructure. Conversely, running a model on company-controlled hardware does not by itself establish that the deployment is secure, compliant, or free of outside dependencies.
Rank #2
- EVOLUTION AMD RYZEN AI MAX+ 395 MINI PC - GMKtec EVO-X2 is the next evolution in AI mini PC Ryzen Strix Halo series. Thanks to AMD Simultaneous Multithreading (SMT) the core-count is effectively doubled, to 32 threads. Ryzen AI Max+ 395 has 64 MB of L3 cache and can boost up to 5.1 GHz, depending on the workload. The Ryzen AI Max+ 395 is currently rated as the "most powerful x86 APU" on the market for AI computing.
- AI NPU with XDNA 2 ARCHITECTURE - Powered by 16 “Zen 5” CPU cores, 50+ peak AI TOPS XDNA 2 NPU and a truly massive integrated GPU driven by 40 AMD RDNA 3.5 CUs, the Ryzen AI MAX+ 395 is a transformative upgrade and delivers a significant performance boost over the competition. The Ryzen AI Max+ 395 excels in consumer AI workloads like the llama.cpp-powered application: LM Studio. Shaping up to be the must-have app for client LLM workloads, LM Studio allows users to locally run the latest language model without any technical knowledge required and unleash their creativity and productivity.
- AMD RADEON 8090S iGPU GAMING PC - The AMD Radeon RX 8060S offers all 40 CUs with up to 2.9 GHz graphics clock and uses the new RDNA 3.5 architecture. The powerful iGPU is positioned between an RTX 4060 and 4070 laptop GPU and therefore enables gaming in FHD at maximum details in most demanding games. The 8060S can also utilize the full 64GB pool, which is perfect for running LLMs such as Deepseek 32B, which runs comfortably on this machine.
- EIGHT CHANNEL LPDDR5X - LPDDR5X is a new ground breaking memory small form factor installed on-board. With blazing speeds up to to 8000MT/s, it runs 1.5x faster than the DDR5 SODIMMs; 90% better performance over DDR5 SODIMMs in video conferencing and photo editing; 30% better performance in productivity apps; 4% better performance in digital content workloads.
- QUAD SCREEN 8K DISPLAY SUPPORT - EVO-X2 AI Mini PC support 4-screen 4K/8K output via HDMI 2.1 (8K@60Hz), DisplayPort 1.4 (4K@60Hz), and dual USB 4 40Gbps Transfer speed (supporting PD3.0/DP1.4/DATA). Ideal for gaming, video editing, and multitasking, it provides expansive and crisp multi-display support.
Three ways companies can deploy AI
| Approach | Data custody and retention | Responsibility and tradeoffs |
|---|---|---|
| Frontier provider with enterprise privacy terms | Depends on the provider’s current terms, selected product, and configuration. Verify retention and data-use terms for the exact deployment. | Can reduce the burden of operating models and infrastructure internally, but leaves the organization dependent on provider policies and availability. |
| Cloud intermediary such as Amazon Bedrock | Fortune describes Bedrock as an entry point intended to let businesses use models without providers seeing their data. That description is Fortune’s, not a universal guarantee; assess the service’s actual data flows and terms. | Can provide access to multiple models through a cloud platform. It still involves a provider relationship and does not remove the need to evaluate security and configuration. |
| Open model on company-controlled infrastructure | Can give the organization more direct control over where data is processed and stored, depending on how the system is built and operated. | Requires more internal expertise, security controls, and ongoing operational work. An open model may not match frontier capabilities for every task. |
Fortune reports that some companies use several models through Bedrock. It also reports that Bedrock customer spend grew 170% in Q1 and that adoption reached nearly 80% of Fortune 100 companies. The story does not specify which Q1 year, its measurement method, or how it defines adoption, so these figures should be read as Fortune’s reported indicators rather than a precise measure of enterprise use.
Why many companies choose a hybrid approach
The deployment choice need not be all-or-nothing. A company might reserve a locally hosted model for a sensitive workflow, use an enterprise cloud service for other tasks, and choose a frontier model where its capabilities are important and the data-sharing terms are acceptable. Fortune describes this kind of hybrid use rather than a broad corporate exit from frontier providers.
The decision is best made task by task. Compare the information involved, the provider’s retention terms, the sensitivity of the output, the model capability the task requires, the infrastructure available, and the staff able to operate the system. A choice that suits an internal, low-risk workflow may be inappropriate for customer records, source code, or regulated information.
Rank #3
- Intel Core Ultra 9 285 Processor: Newly developed cores deliver ultra-smooth and responsive gameplay. AI accelerators prepare users for the next era of gaming on an AI PC.
- Simplistic Design: Enjoy the latest generation of Windows 11 Home for your everyday needs. *MSI recommends Windows 11 Pro for business use.
- NVIDIA GeForce RTX 5070 Ti GPU
- Cool While Gaming: In conjunction with an RGB CPU Air Cooler, the Aegis RS features four system cooling fans; three in the front and one in the rear to pull in cool air and push heat out of the PC.
- Turn on the Bright Lights: With the built-in RGB lighting, take your gaming experience to the next level by pressing the MSI LED button to cycle through lighting options. Customize lighting even further with MSI Center software.
What companies take on when they run open models
More control shifts work and risk inward. Fortune’s sources say open-model deployments demand technical expertise, security controls, and responsible-operation practices; they may also fall short of the most advanced capabilities in areas such as coding or financial analysis. As Gregory put it, “You’re trading control for responsibility, right?”
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
- Security and access: The organization must secure the model environment, control who can reach it, and manage the data and systems connected to it.
- Operations: Staff must provision and maintain the infrastructure, configure the model, and manage updates and failures.
- Capability: The team needs to test whether a particular model is good enough for its actual task; “open” does not mean equivalent to every frontier system.
- Governance: Internal policies and review processes still matter. Self-hosting alone does not make a system secure or compliant.
Can a company run an open model locally?
Yes, some open models can run on company-controlled infrastructure, but the hardware and model must fit the workload. Fortune names a Geekom mini PC as an example capable of running some small local models and explicitly says it cannot handle the most advanced models. The report provides no exact PC model, configuration, or performance test, so that example is not a basis for assuming a particular device will suit a business workload.
Local execution can reduce reliance on a hosted model for some tasks, but the machine is only one part of the system. Data handling, access controls, model setup, and the surrounding applications still determine how information is managed.
How to choose an approach
- Classify the task and data. Identify what information the model will receive, how sensitive it is, and what harm could follow from exposure or an incorrect result.
- Set the required capability. Determine whether the task needs a specific frontier capability or whether a smaller or open model could meet the requirement.
- Check the actual terms and data path. Review current product documentation and contracts for retention, training, access, and where processing occurs. Do not infer these details from a provider’s general privacy statement.
- Compare operational ownership. Decide whether the organization can supply the infrastructure, security expertise, monitoring, and ongoing maintenance a self-hosted deployment needs.
- Pilot the narrowest suitable option. Test with appropriate controls and representative tasks before extending use to more sensitive data or workflows.
Fortune quotes ModMed co-CEO Dan Cane as saying he is “paranoid” about intellectual property but trusts major AI companies to act responsibly because both sides have an interest in security. That view illustrates why organizations may reach different conclusions: risk tolerance, workload, contractual protections, and operational capacity vary.
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.
