South Africa’s first national AI policy draft collapsed on source verification. After fictitious references that appeared AI-generated were found in the bibliography, the government withdrew the document and turned a governance push into a credibility crisis.
Why it matters: enterprise AI coding is moving from individual tools to governed fleets. Databricks says Unity AI Gateway now centralizes controls for Codex, Cursor, Gemini CLI, MCP integrations, budgets, rate limits, and observability.
Stanford HAI’s new report says the measurement gap is now part of the AI story, not a side note. U.S. private AI investment reached $285.9 billion in 2025, while documented AI incidents rose to 362 from 233 a year earlier.
This is less about prestige and more about control of the board. Anthropic says Vas Narasimhan, the Novartis CEO who has overseen more than 35 novel medicines, is joining through the Long-Term Benefit Trust, pushing trust-appointed directors into the majority.
Anthropic updated its Responsible Scaling Policy page on April 2, 2026 and moved the policy to version 3.1. The company says the revision mostly clarifies its AI R&D threshold language and makes explicit that it can pause development even when the RSP does not strictly require it.
OpenAI published a policy blueprint aimed at preventing and combating AI-enabled child sexual exploitation. The framework combines legal modernization, better provider reporting, and safety-by-design measures inside AI systems.
A Hacker News discussion is focusing on a new Linux kernel document that permits AI assistance but keeps DCO, GPL-2.0-only compatibility, and final accountability with human submitters.
Databricks posted on March 27, 2026 that its LogSentinel system uses LLMs to classify columns, apply hierarchical and residency-aware labels, and detect drift, with up to 92% precision and 95% recall for PII on 2,258 samples. Databricks documentation says Unity Catalog Data Classification uses an AI agent and LLM to classify and tag tables, while governed tags and ABAC policies translate those tags into consistent access and compliance controls.
Anthropic said on Mar 11, 2026 that it is launching The Anthropic Institute to study the biggest economic, security, legal, and societal questions raised by frontier AI. The effort is meant to turn observations from inside a model builder into public research and external dialogue.
Microsoft announced Microsoft 365 E7 Frontier Suite on March 9, 2026 as a premium enterprise package that combines Copilot, Agent 365, and advanced security, identity, and compliance controls. The company said the suite will be available on May 1, 2026 for $99 per user per month, alongside a Frontier program that includes Claude and a research preview called Cowork.
Anthropic has published a study on how much autonomy AI agents are being given in the wild using millions of interactions across Claude Code and its public API. The longest Claude Code turns nearly doubled from under 25 minutes to over 45 minutes in three months, while experienced users became more likely to auto-approve and more likely to interrupt when needed.
Anthropic has launched The Anthropic Institute, a new public-interest effort focused on the social challenges posed by powerful AI. The company says the group will combine technical, economic, and social-science expertise to inform the broader public conversation.