The White House is issuing recommendations not to restrict the use of open source AI models, based on a report from the Department of Commerce’s National Telecommunications and Information Administration (NTIA).
The report recommends open models continue to be monitored for potential risks. Part of NTIA’s recommendation is that the government should start an ongoing program for collecting evidence of risks and benefits, evaluating the evidence, and then acting on it.
The report came about because of President Biden’s 2023 Executive Order on “Safe, Secure, and Trustworthy Development and Use of Artificial Intelligence.” One of the things the EO did was instruct NTIA to conduct a report on the “potential benefits, risks, and implications of dual-use foundation models for which the model weights are widely available, as well as policy and regulatory recommendations pertaining to those models.”
Stefano Maffulli, executive director of the Open Source Initiative (OSI) and one of the primary stakeholders of the process, believes these recommendations will be an appropriate way for balancing risks and innovation.
“There are undeniable benefits to research in open models, particularly in transparency around privacy, security and bias,” he said. “We encouraged equitable access to economic benefits by keeping models open, as this is a well-established factor in innovation globally.”
Dr. Thomas Di Giacomo, chief technology and product officer at SUSE, agrees that developing models in the open can help increase trust in AI.
“SUSE believes that all software, including AI, should be freely accessed, used, changed and shared transparently by anyone … We believe community-based standards regarding AI will address the safety concerns of AI, as engaging the developer community is one of the best ways to improve security protocols and resilience,” he said.
You may also like…
Open Source Initiative is close to coming up with a definition for Open Source AI
Open source in 2024: Tackling challenges related to security, AI, and long-term sustainability
