This week's AI policy update.

First, new models, new alternatives

This week, OpenAI released two open-weight AI models, marking a potential change in strategy for the developer of ChatGPT.

Commenting on the release, Maximillian Gahntz, AI policy lead at the Mozilla Foundation, asks whether it's fully open-source, why OpenAI has done this, and if it's good, bad or in between.

For policymakers, he suggests governments must continue to invest in open-source AI and cultivate alternatives developed with the public interest and the cultural richness of the entire world in mind.

See Maximilian's insightful post here: https://lnkd.in/ei5GHpRc

Second, new blog: AI policy in practice

Kicking off a new series of articles for The Appraise Network, we spoke to Shradha N Mathur, of AI start-up OpenSphere about why AI policy isn't abstract and the need for greater participation in the policy process.

As Shradha states: "Good AI policy starts with real-world context. If we want systems that work for people, we need rules shaped by the people they affect."

Read on to learn more about how Shradha helps ensure responsible AI development and deployment, and her views on how policymakers might improve AI regulation.

See Shradha’s article here: https://appraisenetwork.ai/ai-policy-in-practice-shradha-mathur

Last, new course: the EU AI Act for builders.

The excellent Merve Hickok, President of the Center for AI and Digital Policy, has launched a new LinkedIn course on the EU AI Act.

The course covers the objectives behind the regulation, how to determine whether a system may fall under the Act, and the implications of the Act for various AI applications.

See Merve's post here for more: https://lnkd.in/e2jGc5pu

As always, keen to hear what others are reading.