As tech giants like Microsoft and Meta push policymakers to support open-source AI systems, they’re arguing that this approach is a safer bet than relying on proprietary models. In an open letter posted last week, over two dozen companies make the case that open-source AI will be more beneficial in the long run for several reasons.
The analogy drawn by these companies is with the software industry of the 1980s. Back then, large businesses were initially concerned that open-source code would cut into their profits. However, the end result was a thriving ecosystem that underpins much of modern computing, government IT, and even commercial software products today. This shared foundation of knowledge has also fueled countless future innovations.
The letter’s authors argue that the US is at a similar crossroads with AI. Instead of relying on a single proprietary model, they say the country should focus on building an open ecosystem that can be adapted to various sectors. This approach would allow startups, universities, and research labs to innovate and iterate on AI technology, making it more broadly useful to society.
However, expanding access to open-source AI also comes with significant security risks. Cybersecurity experts warn that low-level criminals may exploit these tools to launch serious attacks. Once an AI model is made available, anyone can download and customize it for malicious purposes. This has already led to the proliferation of deepfakes and other AI-generated imagery, which could be used to create locally customized child exploitation materials (CSAM) or sexualized deepfakes.
Despite these concerns, the letter’s authors believe that open-source AI is more beneficial in the long run. They argue that this approach would allow organizations to match the right model to the right job at the right cost, reserving frontier-scale capability for genuine frontier problems. This discipline is what will make AI economically sustainable as it scales into everyday tasks.
For cybersecurity specifically, the letter’s authors believe that defenders armed with open-source AI will outpace attackers better than any closed model approach. They argue that open models broaden defensive capabilities, increase transparency, and allow vulnerabilities to be discovered and remediated across many teams.
The letter has sparked debate about how policymakers should balance support for the domestic AI industry with oversight and regulation of potential harms. The Trump administration has cycled through several frameworks since taking office, from a laissez-faire approach to creating voluntary testing regimes and imposing export controls on specific models.
As US companies continue to navigate this complex landscape, it’s clear that there are no easy answers. However, by supporting open-source AI, policymakers may be able to strike the right balance between innovation and regulation, ultimately making society safer from cyber threats. One practical takeaway for readers is to stay informed about the latest developments in AI policy and consider participating in online discussions about how these technologies can be developed responsibly.
Source: CyberScoop — 2026-07-24