AI systems pose black-box decision-making challenges, leaving researchers, users, and regulators grappling with a technology they can’t fully observe. This brief reviews the history, current state, and best practices of algorithm auditing.
POLICY
-
Cloud Companies Improving Open Source Support and Licensing
By
–
Years later, some sw companies have gotten better at operating it and AWS (among other clouds) has gotten better at contributing to the sw. But there's still a gap and that prompts things like the BSL.
-
Algorithmic Bias in Financial Decision-Making Systems
By
–
A partial description from the #SXSW pitch: "Predictive, data-driven algorithms are increasingly used to make life-changing decisions—including who gets access to financial services. But are algorithms the unbiased actors we’re led to believe?" #SXSW2024
-
Building Responsible AI Models: Inclusion, Bias, Investment
By
–
The proposed @SXSW discussion will focus on building responsible #AI models that drive inclusion, avoiding algorithmic bias & discrimination, using investor tools for vetting AI startups and more. #SXSW #SXSW2024
-
AI Risks and Rewards in Financial Services Panel at SXSW
By
–
I've been invited to moderate a @sxsw 2024 panel to about potential risks & rewards of AI in financial services — an important, relevant & complex topic applies to every industry. #SXSW Vote for our panel before 8/20 via the #SXSW2024 PanelPicker page: https://
shar.es/afNfBX -
AI, Justice and Safety: Ada Lovelace Institute Leadership
By
–
The interconnectedness of AI, justice and safety requires a forward looking and holistic vision for sociotechnical research, community engagement and policy insights, and I’m excited to join the board of the Ada Lovelace Institute to help strengthen its leadership and impact
-
Open-source debate: fear tactics and technology autonomy
By
–
I don't want to imply bad faith but this tweet really makes it look like you're using fear to fight and discredit open-source. On your actual point, I'm not really sure what "survive and spread on their own" means for technology but if it can do that, isn't creation itself
-
Nerd Joins Podcast With Policy Scholars and Economists
By
–
Niall @nfergus calls me a “nerd”😜. So a nerd walks into a podcast hosted by three historians, economists and policy scholars @HooverInst @StanfordHAI @LTGHRMcMaster @JohnHCochrane … 😜 https://t.co/sVR2B3hwxo
— Fei-Fei Li (@drfeifei) 10 août 2023Niall @nfergus calls me a “nerd”. So a nerd walks into a podcast hosted by three historians, economists and policy scholars @HooverInst @StanfordHAI @LTGHRMcMaster @JohnHCochrane …
-
Stanford scholars recommend fairness training for decision-making algorithms
By
–
Algorithmic decisions based on fairness may not always lead to equitable or desirable outcomes. In a new paper, HAI faculty affiliate @JulianNyarko and other Stanford scholars offer recommendations for training algorithms to assist in decision making.
-
Congressional AI Boot Camp Year 2 Concludes with Policy Resources
By
–
That’s a wrap on our 2nd year of the congressional boot camp on AI! Thanks to all the dedicated congressional staffers and the high caliber of speakers who joined us over the past 3 days. For more resources on AI policy, check out our publications here: https://
hai.stanford.edu/policy/policy-
publications
…