#ai-policy — Public Fediverse posts
Live and recent posts from across the Fediverse tagged #ai-policy, aggregated by home.social.
-
🥱 Wow, a groundbreaking revelation that efficient workflows might be beneficial to AI policy. Who knew?! 🤔 And by all means, let's solve complex AI issues the same way we managed foreclosure paperwork a decade ago. Because nothing says cutting-edge technology like 2008 law firm vibes. 📜🔨
https://brianmeeker.me/2026/05/14/have-a-coherent-ai-policy/ #efficientworkflows #AIpolicy #groundbreakingrevelation #legaltech #2008vibes #HackerNews #ngated -
Anthropic publishes exact API rates and plan prices down to the dollar, but keeps subscription token limits deliberately undefined—and adjusts them without notice. The pattern suggests limits are a pricing lever, not a fixed product feature. What changes next month?
-
South Africa Used AI To Write Its Now Withdrawn AI Policy. The Citations Were Fake.
-
Many Americans pessimistic about AI's impact—and want more regulation
https://phys.org/news/2026-05-americans-pessimistic-ai-impact.html
-
In The Vacuum Of AI Legislation, Libraries Have The Playbook
-
AI digest covers local model tradeoffs on M4 hardware, GrapheneOS warnings on remote attestation as a computing kill switch, and Anthropics explanation for Claude attempting to blackmail engineers during testing.
-
Frontier LLMs corrupt 25% of documents in long workflows per new benchmark, while a Fields Medalist reports ChatGPT 5.5 Pro solving PhD-level math. Mayo Clinic AI detects pancreatic cancer years early.
-
@marcia You can have a look at #Concordia’s teaching guidelines re #genAI —
https://www.concordia.ca/ctl/tech-tools/teach-with-technology/guidelines-gen-ai.htmlThere was also a project on genAI in #highered that involved Concordia —
https://www.concordia.ca/news/stories/2023/11/16/concordia-joins-two-year-collaborative-research-project-making-ai-generative-for-higher-education.htmlI have started looking for the reports mentioned here (by the project lead), but no success as of yet —
https://sr.ithaka.org/blog/making-ai-generative-for-higher-education-2/ -
OpenAI debuts GPT-5.5 Instant with fewer hallucinations as the new ChatGPT default, while reports surface of an OpenAI smartphone targeting 2027 and Meta faces a new publisher lawsuit over training data.
-
Today the #SROP webmaster made the first update to our FAQ page since 2023. Reluctantly, we have added an "AI Policy" to clarify things that none of our myriad customers have, so far, asked about. We might elaborate at some point, but at least now we have a "policy".
Any morbidly curious Mastodonians can follow the link here:
-
Musk v. Altman trial enters week two with Brockman on the stand and damaging journal entries surfacing. Sierra raises 1B for enterprise AI agents, Cerebras files for a major IPO, and llama.cpp may double local inference speeds.
-
By early 2026, around half of US jobs had at least a quarter of their tasks done with help from AI (Anthropic Economic Index). Klarna replaced 700 customer service agents with AI in 2024, then began rehiring after customer satisfaction dropped.
A new piece on 10Billion asks whether upskilling has become exhaustion dressed up as ambition, and what a real menu of structural responses would look like.
https://10billion.substack.com/p/the-finish-line-keeps-moving
-
DeepSeek V4 and Kimi K2.6 challenge frontier labs on price and benchmarks, a Hawley bill to ID-verify chatbot users clears committee, and Microsoft quietly attributed commits to Copilot without its use.
-
Pentagon expands classified AI vendors while excluding Anthropic over weapons restrictions, xAI launches Grok 4.3 at aggressive pricing, and Meta cites AI spending in laying off 8000 workers.
-
Washington insiders admit AI is slipping beyond government control
https://fed.brid.gy/r/https://nerds.xyz/2026/05/ai-regulation-policy-insiders-warning/
-
🇵🇪 Peru just published its AI National Strategy for the 2026-2030 period.
-
Google signs classified Pentagon AI deal as Musk testifies against Altman in court. OpenAI models also hit AWS hours after Microsoft deal closed, and a 13B model was trained solely on pre-1931 text.
