The Daily AI Briefing

The Daily AI Briefing — 3 June 2026

AI news5

  1. Microsoft launches "Autopilots" and Scout at Build 2026

    At Build on 2 June, Microsoft introduced Autopilots, a class of always-on AI agents that run in the background with their own identity and act without being prompted each time. The first, Microsoft Scout, is powered by the open-source OpenClaw platform and works across Teams, Outlook, OneDrive and SharePoint. It is out as an early experimental release to Frontier enterprise customers, alongside a previewed Copilot super app and a Fireworks AI partnership for Azure AI Foundry.

    Why it matters

    Always-on autonomous agents shift AI from something you prompt to something that works in the background, the model most enterprise software will follow.

    microsoft.com
  2. Trump signs a narrower AI oversight executive order after industry pushback

    On 2 June, President Trump signed an executive order asking AI companies to voluntarily submit new models to the government for evaluation 30 days before public release. It also directs federal agencies to build cyber-capability benchmarks and an "AI cybersecurity clearinghouse" to share vulnerability information. The order was narrowed from an earlier 90-day version after objections from industry, including former White House AI czar David Sacks.

    Why it matters

    It signals the US will lean on light-touch voluntary review rather than hard regulation, with competition from China kept front of mind.

    techcrunch.com
  3. Anthropic scales Claude Mythos to critical infrastructure across 15-plus countries

    Anthropic expanded Project Glasswing, giving 150 more organisations across more than 15 countries access to Claude Mythos, its most powerful security model, to scan for vulnerabilities. The expansion targets critical infrastructure in power, water, healthcare and communications. Named recipients include Okta, Samsung, SK Hynix, SK Telecom, NATO and the EU cyber agency ENISA. Glasswing partners have so far found more than 10,000 high or critical-severity flaws.

    Why it matters

    Frontier AI is moving from finding software bugs to defending the systems nations depend on, a high-stakes new front for AI in security.

    techcrunch.com
  4. Uber caps employee AI spending after blowing its budget in four months

    Uber has capped employee AI spending at $1,500 a month per agentic coding tool, such as Claude Code and Cursor, after exhausting its annual AI budget in four months. The CTO said in April that individual engineers were running up bills of $500 to $2,000 a month in token use. Spending is now tracked on an internal dashboard, and caps can be exceeded with permission.

    Why it matters

    If even Uber cannot absorb open-ended AI tool costs, it is a real-world signal of how fast usage-based AI bills escalate for everyone.

    techcrunch.com
  5. Google rolls out fake-call detection to counter AI voice-cloning scams

    Google is rolling out fake-call detection in Phone by Google to Android 12+ devices this month, starting with Pixel. When both the caller and recipient use the app, their phones exchange a silent end-to-end encrypted RCS signal to confirm the call genuinely comes from that contact, countering AI deepfake impersonation. INTERPOL has tied impersonation fraud to more than $400 billion in global losses.

    Why it matters

    As voice deepfakes get cheap, device-level verification of who is really calling becomes basic safety infrastructure.

    techcrunch.com

AI in the nonprofit sector1

  1. Humanity AI preparing a $10m open call for public-interest AI

    Humanity AI, the $500m coalition of ten foundations including MacArthur, Ford, Mellon, Omidyar and Packard, is preparing a $10 million open call, expected to open in the coming months, for organisations working at the frontier of AI and the public interest. It follows the coalition's first $8 million in grants to nine groups (announced 12 May) spanning democracy, labour rights, journalism and education. The open call is designed to widen access beyond the inaugural, invited grantees.

    Why it matters

    It is one of the largest open funding rounds aimed squarely at public-interest AI, worth flagging for nonprofits building in the space.

    humanityai.ai

Free, and actually free

Courses, templates and working tools for nonprofits

No card, no trial. The Briefing publishes here every morning.

Open the Hub →