In the United States, the Biden-Harris Administration has hailed voluntary commitments from seven big tech companies to move toward safe, secure, and transparent development of AI technology. The White House said it is developing an executive order and will pursue bipartisan legislation on ‘responsible innovation’.
Amazon, Anthropic, Google, Inflection, Meta, Microsoft, and OpenAI commit among other things to internal and external security testing of their AI systems before release; sharing information across the industry and with governments, civil society, and academia on managing AI risks; facilitating third-party discovery and reporting of vulnerabilities in their AI; and publicly reporting their AI systems’ capabilities, limitations, and areas of appropriate and inappropriate use.
See a post on the Anthropic website on the AI developer’s approach to “frontier threats red teaming,” and findings from a project on biological risks as a test. As for how wide are the potential effects of artificial intelligence, Inflection has blogged about the 2024 presidential election in the US, ‘and it will be the first in the era of generative AI. Preserving the stability and integrity of our electoral system means starting work right now.’
Separately, see the open letter on the website of the World Ethical Data Foundation, offering ‘first questions and considerations we believe every AI team and individual builder should take everyday to ensure we are releasing more ethical AI models’.
Comment
Michael Covington, VP of Strategy at the device management access product company Jamf, said: “It is encouraging to see an emerging consensus from leading AI innovators as the community takes steps to more effectively manage the risks that accompany early AI adoption. ChatGPT and other examples of generative AI have experienced massive growth in an environment that had near-zero guardrails in place. There is value in making sure that organisations properly assess their AI developments so this truly innovative technology can develop in a more controlled fashion.
“It is essential to ensure that developers and users alike are aware of the benefits and dangers that accompany AI. At Jamf, we believe in empowering users and giving them more control and authority when it comes to decision-making. The more stringent testing, research, and transparency that will come out of this announcement helps ensure that the power of AI is thoughtfully harnessed in its future applications.
“The companies developing AI — from Amazon to Tesla — all will face challenges with public perception as they manage through the highs and lows of how this new technology is released into the market. Innovation is wonderful when it is fully realised; but all innovation involves risk and requires patience through failure. Managing expectations on what capabilities AI can deliver, and providing guidelines for both developers and consumers is critical for the technology’s long-term success to be realised.”
