Recently, AI company Anthropic quietly removed from its website commitments to AI safety made during the Biden administration. These commitments, initially discovered by an AI oversight group called the "Midas Project," were deleted last week from Anthropic's transparency center, which lists the company's "voluntary commitments" to responsible AI development. While not legally binding, these commitments pledged to share information and research on AI risks (including bias) with the government.
In July 2023, Anthropic, along with other major tech companies including OpenAI, Google, and Meta, joined the Biden administration's self-regulatory voluntary agreement to support AI safety initiatives. These initiatives were further solidified in Biden's AI executive order. Participating companies pledged to conduct safety testing on models before release, watermark AI-generated content, and develop data privacy infrastructure.
However, with the Trump administration's arrival, Anthropic's stance on these commitments appears to have shifted. On his first day in office, Trump rescinded Biden's executive order, dismissed several government AI experts, and cut research funding. These changes may have prompted many large AI companies to reassess their relationships with the government, with some potentially seeking to expand government contracts to influence the still-unclear AI policy landscape.
Anthropic has not publicly commented on the removal of the commitment, stating its position on responsible AI is unrelated to, or predates, the Biden-era agreements. Relatedly, the Trump administration is likely to dismantle AI safety institutes established during the Biden era, leaving related initiatives uncertain.
Overall, the Trump administration is weakening the AI regulatory framework established under Biden, leaving AI companies with seemingly greater freedom to manage their systems without external regulatory pressure. Currently, safety checks regarding AI bias and discrimination do not appear to be a focus of the Trump administration's relevant policies.
Key Takeaways:
🌐1. Anthropic's removal of AI safety commitments made during the Biden administration reflects a policy shift.
🤖2. The Trump administration has rescinded Biden's AI executive order and is cutting back on AI regulation.
💼3. Large AI companies are reassessing their relationships with the government under the Trump administration, potentially leading to relaxed self-regulation.
