Natural language processing

Meta accused of “open washing” AI models, clashing with open-source purists

Meta accused of “open washing” AI models, clashing with open-source purists



summary
Summary

Meta has been criticized by the open-source community for its approach to AI development, as the company seeks to define open-source AI on its own terms while potentially exploiting regulatory loopholes.

Meta CEO Mark Zuckerberg is facing accusations of “open washing” the company’s AI models, as the tech giant clashes with open-source purists over the definition of open-source artificial intelligence.

The Open Source Initiative (OSI) recently issued draft standards for open-source AI, which Meta’s popular Llama models fail to meet. The OSI’s definition requires developers to make available sufficient information about training data, source code, and internal model weights to enable replication.

Meta releases the weights of its Llama models but withholds training data and imposes licensing restrictions, falling short of the OSI’s criteria.

Ad

Stefano Maffulli, head of the OSI, accuses Zuckerberg of “really bullying the industry to follow his lead” in defining open-source AI, The Economist reports.

The publication also quotes Ali Farhadi, director of the Allen Institute for AI, which developed the more transparent OLMo model. He acknowledges Llama’s contributions, but says, “We love them, we celebrate them, we cherish them. They are stepping in the right direction. But they are just not open source.”

Critics argue that Meta’s approach may be an attempt to exploit regulatory loopholes. The EU AI Act, which became law this year, offers exceptions for open-source models. However, the legislation contains conflicting definitions of what constitutes open-source AI, notes Kai Zenner, a policy adviser at the European Parliament, as reported by The Economist.

The Economist cites Mark Surman, head of the Mozilla Foundation, who warns of the risk of “open-washing” without a precise definition of open-source AI. He argues that a clear definition would give developers confidence in using, copying, and modifying models like Llama without being “at the whim” of Zuckerberg’s goodwill.

Meta’s Defense and Industry Realities

Meta defends its position, objecting to the OSI’s binary approach and arguing that the cost and complexity of developing large language models (LLMs) requires a spectrum of openness in which developers decide for themselves how to release their models. The company claims that few models meet the OSI’s definition, none of which are state-of-the-art.

Recommendation

The debate over the definition of open-source AI gains additional significance as regulations evolve. California’s SB 1047 bill, aimed at responsible AI development in Silicon Valley, has prompted open-source advocates to urge for a precise definition of open-source AI in collaboration with the OSI.

Meta seems to be strategically trying to reap the benefits of working with the open-source community, such as attracting developers to its AI infrastructure, while keeping its treasure trove of training data to itself. A few weeks ago, Zuckerberg warned about the risk of over-regulation in Europe in an essay with Spotify CEO Daniel Elk.

Meta accused of "open washing" AI models, clashing with open-source purists

Source link