AI Regulation Updates, 10 October 2024

The H.A.I.R AI Regulation Updates are sponsored by Warden AI.

Warden AI is the tech-led AI auditor that helps HR Tech platforms and HR/TA teams demonstrate AI fairness and compliance. Our platform continuously audits AI systems for bias and provides transparency with real-time dashboards and reports. We work with leading AI talent platforms like Beamery and enterprise HR teams to mitigate risk, build trust, and meet AI regulations like NYC LL 144 and the EU AI Act.

Where 8 AI frameworks find common ground

The nonprofit Partnership on AI released a policy paper looking at the interoperability challenges and opportunities between current major policy frameworks for foundation models. The report emphasizes the need for global coordination in creating policy frameworks for foundation models in AI, particularly focusing on transparency and documentation requirements. For recruitment and HR, the key takeaway is that consistent documentation and transparency across different AI governance frameworks will enhance accountability and safety.

World Economic Forum publishes report on AI

The World Economic Forum published a report on AI governance outlining how organizations and governments can enforce AI safety measures. The WEF said the report provides "policy-makers and regulators with implementable strategies for resilient generative AI governance contained within a 360-degree governance framework."

Time to show AI results, company leaders say

This WSJ article discusses the shift from experimental AI projects to focusing on measurable returns on investment (ROI). For recruitment and HR professionals, this means AI initiatives must now deliver clear business outcomes, like improved productivity or revenue, within 12 months. Many generative AI projects are still in the testing phase, and companies are realizing AI isn't a "magic wand" but a tool that must be integrated thoughtfully into business strategies. HR leaders should focus on proven AI solutions that enhance recruitment efficiency, automate repetitive tasks, or improve candidate experiences, while ensuring these technologies deliver tangible benefits to the business.

EU general-purpose AI Code of Practice timeline takes shape

Euractiv reports a first draft of the EU Code of Practice for general-purpose AI could be ready 3 Nov., while a European Commission spokesperson indicated a final draft is expected April 2025. The timeline comes after the Commission's consultation on the code received approximately 430 stakeholder comments. The Commission also disclosed disagreements that surfaced between general-purpose AI model developers and various stakeholders during first plenary session for code drafting.

How to place civil rights protections into privacy, AI standards

The Leadership Conference on Civil and Human Rights released two reports examining the path forward for civil rights privacy and AI safeguards based on foundational civil rights protections. The coalition argues such protections are critical to ensure people's civil rights are upheld and data is not misused to perpetuate bias.

FCC fines man at the center of AI-generated robocalls

The U.S. Federal Communications Commission issued a USD6 million fine to the person who allegedly made deceptive robocalls to New Hampshire voters using generative AI technology. The caller pretended to be President Joe Biden and spread election misinformation, a violation of the Truth in Caller ID Act.

AI assistant's impact on workplace privacy

AI workplace tools, such as Otter.ai and Zoom’s AI Companion, are raising privacy concerns by inadvertently sharing sensitive information. These tools can capture private conversations, even after meetings end, leading to breaches of confidentiality. In recruitment and HR, this is particularly concerning for sensitive discussions like layoffs or evaluations. Both companies and employees need to carefully manage AI settings to prevent unintended sharing. Employers should also ensure that AI tools are implemented responsibly and users are educated about the potential risks to avoid privacy mishaps.