Ecosystem Ramps up Open-Source and Open-Standards AI Investment as Vendors Look to Monetize Long-Term Market Trend

The open-source AI market has been buzzing for a while, but increasingly, it has become clear that it will dominate both the Generative Artificial Intelligence (Gen AI) and traditional AI markets in the long term. This bullish outlook is underpinned by 1) historic trends in software markets, especially within traditional AI (computer vision, Natural Language Processing (NLP), and predictive AI), and 2) its long-term enterprise value proposition. Although proprietary or closed-source AI still dominates enterprise Proofs of Concept (PoCs), it is expected that as the enterprise market scales, more will transition toward open-source AI. This is driven by multiple commercial and technology factors—transparency, innovation, developer friendliness, long-term cost profile, growing Application Programming Interface (API) and cloud hosting costs, security, ownership, control, vendor lock-in, and other factors. These factors will push enterprises to spend more on software to develop and deploy open-source AI. ABI Research forecasts that, by 2030, around US$250 billion will be spent on tools and platforms to support open-source AI deployment—making up around 65% of total AI software spending. The supply side clearly agrees, given the scope of the open source market activity this year:

1) Gen AI leaders increasingly ramp up Research and Development (R&D) and investment in open source:

• Google, which previously focused investment in closed models like Gemini and Bard, recently released Gemma, a lightweight open-source model, and Project Oscar, an open-source framework to create AI agents.
• AI21 complemented Jurassic-1 (closed) with Jamba, an open model family.
• Apple AI released the open-source 7B language model trained on open datasets aiming to accelerate R&D.
• Meta placed itself at the center of the open-source AI market with massive investment in NVIDIA hardware, the release of models that rival closed-source alternatives, and long-term pledges to invest in R&D.

2) Open-source developers and contributors are snapped up by AI vendors:

• Intel acquired Granulate to develop Tiber, an open-source platform supporting application optimization for the cloud and on-premises.
• AMD recently acquired Nod.ai to enhance its open-source software capabilities.
• AI Squared acquired Multiwoven to integrate open-source Reverse ETL into an enterprise offering.
• Protect AI acquires Laiyer AI to integrate LLM Guard with its open-source Large Language Model (LLM) security capabilities.
• IBM announces intention to acquire HashiCorp, the developer and maintainer of Terraform.
• DataStax acquires Langflow to support customers with quickly building out new AI stacks.

The supply side is obviously bullish around the enterprise opportunity within open-source AI software. In addition, open standards are increasingly being invested in as the ecosystem looks to break NVIDIA’s dominance. Unified Acceleration Foundation (UXL), built on top of Intel’s OneAPI, is being developed to drive an open-standard accelerator ecosystem; Ultra Accelerator Link (UALink), supported by AMD and other industry leaders, aims to challenge NVLink by developing an open solution to scale up networking for AI accelerator systems.

Open-source AI investment is expensive and challenging, and for cash-light companies without a clear long-term monetization strategy, it can bring significant risks. Many question why they should invest millions or billions into models, tools, or datasets that will be made freely available to anyone. However, open-source investment, if done strategically, can create enormous long-term value.

Below, ABI Research highlights some of the strategies that sit behind open-source investment:

• Building Brand Recognition, Community, and User Base: Open-source AI is a community project that enables a range of stakeholders to contribute and maintain projects. Certain vendors will invest in open source to build a community-based platform with traction. Often, this will be the first step to be followed by monetization. Hugging Face is an example of a company that has built a strong brand with open source, but has yet to monetize this “future” customer base.
• Alternative to Freemium: Probably the most common strategy, open-source is a strong alternative to freemium with the added advantage that the product benefits from cross-ecosystem innovation. GitHub (recently acquired by Microsoft) is an example of a company that is starting to transition its strategy from pure community to a “freemium” model through GitHub Copilot for enterprise.
• Development of Ecosystem Value Proposition: A strategy deployed by data and AI platforms and chip vendors, among others. Investing in open-source AI will drive new users toward platforms to provide access to optimized solutions within specific ecosystems. Intel invests in open source to build solution differentiation, and drive customers into its ecosystem.
• Scorched Earth Approach: Undermining existing market solutions to drive growth of a customer base for other products. Meta is certainly leveraging this strategy to ensure competitors cannot build moats with closed-source tooling. This will help Meta dominate the market and integrate products into its family of applications. A scorched Earth approach will work best for large companies that have other features creating differentiation—in Meta’s case, this is its application ecosystem and huge user base.
• Low-Cost R&D with Focus on Commercialization: AI is expensive, and continuously investing in new models is challenging. Open-sourcing can enable different stakeholders to invest in this process, with the maintainers potentially reaping the benefits through service/solution differentiation. This is probably the most widely deployed open-source strategy.

Given supply side investment in open source, and the long-term preferential value proposition for enterprises, ABI Research is very bullish on this opportunity. However, building an effective open-source AI strategy is challenging, especially in the context of a hyper-competitive AI space. So, as the market trends toward open source, this is where incumbents focus to remain relevant and maximize value created from open-source AI:

• Establish Governance: This should be the first step for any company building open-source AI. It is vital to set a clear vision with goals and a leadership structure. This is especially important given that most projects are led by two major companies that define the roadmap, community conduct, licensing, and guidelines.
• Develop Short, Medium, and Long-Term Open-Source Strategy: This should address expectations for maintenance, contribution, monetization, and costs. Open-source R&D is expensive and brings very few short-run opportunities, so stakeholders must think over a 5+ year timeframe.
• Assess Developer Capabilities against Open-Source Expectations: Contribution and maintenance is time- and skill-intensive, so vendors looking to build a strategy here must assess internal developer resources and evaluate feasibility.
• Assess Competitor Landscape to Understand Options: Given the nature of open source, developer projects may already be in progress and internal investment would be a waste.
• Set a Clear Divide between Contributions and Monetization: Open-source R&D must be carefully calculated with a clear divide between contribution and in-house differentiation that can be monetized.
• Ensure Developers Remain Closely Aligned with Internal Strategy: Open-source contributions/projects can often be misaligned with corporate open-source strategy, ensuring that developers are closely monitored and assessed against Key Performance Indicators (KPIs) will increase opportunities for effective direct or indirect open-source monetization.
• Build Partnerships to Coordinate Contributions and Share R&D Costs: Open-source strategies do not need to be internal; in fact, more collaboration between partners will improve time-to-value, market traction, brand recognition, and other factors.
• Assess Mergers and Acquisitions (M&A) Opportunity to Bring Open-Source Developers In-House: Acquiring small teams will help accelerate efforts and bring open-source projects in-house, which may have commercial value. As we have seen, open-source players have been bought for a huge valuation, but this may be worth it, given the long-term enterprise opportunity in open source.

Effective open-source strategies will certainly pay dividends over the long term, as it helps build platform stickiness, brand recognition, and may directly support monetization. But this is costly and very challenging, especially for those investing in leading-edge model development. Even open-source stalwarts like Intel have struggled to leverage their significant investment in open source over the last few years to effectively monetize AI—especially Gen AI. Ensuring that companies don’t struggle in the same ways means focusing on building strong internal processes that drive alignment all the way from developers to the C-suite. Investment in open source is risky, but even more risky would be for leaders to ignore this trend that will dominate the Gen AI market moving forward.