Senators Query Meta's CEO Over LLaMA AI Model -

Join Our Telegram channel to stay up to date on breaking news coverage

The Meta CEO, Mark Zuckerberg, is in the crosshairs of two U.S. Senators. Senator Richard Blumenthal (D-CT), who is chair of the Senate’s Subcommittee on Privacy, Technology, & the Law, and Josh Hawley (R-MO) have sent a letter to the Meta CEO, raising concerns about the leak of Meta’s groundbreaking open-source large language model LLaMA.

Meta released its advanced AI model, LLaMA, w/seemingly little consideration & safeguards against misuse—a real risk of fraud, privacy intrusions & cybercrime. Sen. Hawley & I are writing to Meta on the steps being taken to assess & prevent the abuse of LLaMA & other AI models. pic.twitter.com/vDyJbuWSlJ

— Richard Blumenthal (@SenBlumenthal) June 6, 2023

The senators have expressed their concern over the “potential for its misuse in spam, fraud, malware, privacy violations, harassment, and other wrongdoings.” The two wrote:

We are writing to request information on how your company assessed the risk of releasing LLaMA. What steps were taken to prevent the abuse of the model, and how you are updating your policies and practices based on its unrestrained availability.

The senators further accused Meta of “doing minimal” to censor the model. The senators said:

When asked to write a note pretending to be someone’s son asking for money to get out of a difficult situation, OpenAI’s ChatGPT will deny the request based on its ethical guidelines. In contrast, LLaMA will produce the requested letter and other answers involving self-harm, crime, and antisemitism.

The LLaMA model

LLaMA is one of the most extensive open-source Large Language Models currently available. Nonetheless, most of the popular uncensored LLMs shared today are LLaMA-based. This reaffirms LLaMA’s central role in LLM’s sphere. For an open-source, it is incredibly sophisticated and accurate.

For example, Stanford’s Alpaca open-source LLM, released mid-March, utilizes LLaMA’s weights. Vicuna, a fine-tuned version of LLaMA, matches GPT-4’s performance, further attesting to LLaMA’s influential role in the LLM sphere. This indicates that LLaMA plays a vital role in the current status of open-sourced LLMs.

Notably, LLaMA’s release took effect in February. The senator’s letter, however, points to LLaMA’s release, asserting that Meta released LLaMA for download by approved researchers “rather than centralizing and restricting access to the underlying data, software, and model.”

The controversy arises from the open dissemination of LLaMA. Shortly after its announcement, the whole model surfaced on BitTorrent, rendering it accessible to anyone, anywhere in the world, without monitoring or oversight. The open dissemination of LLaMA represents a significant increase in the sophistication of the AI models available to the general public. It raises serious questions about the potential for misuse or abuse.

Senators Criticize Meta’s Use of Word “Leak.”

The Senators had harsh words for Zuckerberg regarding LLaMA’s distribution and the use of the word “leak.” However, the focus on the issue arises at a time when new and advanced open-source language AI developments launched by startups, academics, and collectives, are flooding the internet.

The letter charges that Meta should have foreseen the broad dissemination and potential for abuse of LLaMA, given its minimal release protections. The letter asserts:

Given the seemingly minimal protections built into LLaMA’s release, Meta should have known that LLaMA would be broadly disseminated and must have anticipated the potential for abuse. While Meta has described the release as a leak, its chief AI scientist has stated that open models are critical to its commercial success. Unfortunately, Meta appears to have failed to conduct any meaningful risk assessment before release, despite the realistic potential for broad distribution, even if unauthorized.

Additionally, Meta has made LLaMA’s weights available case-by-case for academics and researchers, including Stanford, for the Alpaca project. However, these weights subsequently leaked, enabling global access to a GPT-level LLM for the first time. Model weights are a component of LLMs and other machine-learning models. At the same time, an LLM is a specific instance that uses those weights to provide a result.

It is important to note that none of these open-source LLMs are available yet for commercial use. This is because the LLaMA model is not released for commercial use, and the OpenAI GPT-3.5 terms of service prohibit using the model to develop AI models that compete with OpenAI.

As the debate rages on open-source AI models’ risks and benefits, the dance between innovation and risk continues. All eyes in the open-source LLM community remain firmly on the unfolding LLaMA saga.

Newest Meme Coin ICO - Wall Street Pepe

Rating