Keldura Daily Open Keldura

Keldura Daily · Technology and AI

Technology and AI: The most actionable developments

This set focuses on concrete shifts in AI platform governance, enterprise AI product positioning, compute infrastructure deals, and policy pressure around how AI is used in the real world. Several items are still reported as leaks, talks, or enforcement actions rather than completed launches, so the emphasis is on what has actually changed and what observers can verify next.

The field note

2 sources · 3 items
  1. Apple says it removed three of the apps named in the demand and is terminating the developers’ accounts from it…
  2. The company framed the apps as violations of existing App Review Guidelines rather than a new policy category,…
  3. User reporting tools were part of the removal path, which means platform moderation is not relying only on exte…
Story 011 source

Apple is actively removing AI “nudify” apps and terminating developer accounts

Apple said it removed three apps from the App Store after San Francisco city attorney David Chiu demanded that Apple and Google take down 13 apps capable of generating non-consensual nude images [1]. Apple also said it has “strictly prohibited” apps designed to generate, distribute, or consume pornography, and that it has proactively rejected many “nudification” apps and removed others when users flagged them [1].

Why it matters

This is a direct example of platform enforcement against abusive generative-AI use, and it shows that app-store policy is becoming a real choke point for harmful AI products [1]. The action also signals that legal pressure from city officials can move quickly into product takedowns and developer-account sanctions [1].

Key insights

  • Apple says it removed three of the apps named in the demand and is terminating the developers’ accounts from its program [1].
  • The company framed the apps as violations of existing App Review Guidelines rather than a new policy category, suggesting enforcement is being pushed through current rules [1].
  • User reporting tools were part of the removal path, which means platform moderation is not relying only on external legal complaints [1].
Story 021 source

MLB is banning generative-AI apps on dugout iPads used for in-game decisions

Major League Baseball is cracking down on league-issued iPads that teams have used since 2016 by banning custom apps that tap generative AI for “substitutions, pitch calling, and other in-game decisions traditionally made by players and coaches” [10]. The report says as much as a third of the league used the dugout iPads for at least one of those purposes [10].

Why it matters

This shows sports leagues drawing a line between analytics tools and AI systems that directly influence live decisions [10]. It is also a useful signal for other organizations that issue controlled devices: if AI is being used in high-stakes decision workflows, expect tighter restrictions and more explicit governance [10].

Key insights

  • The ban applies to custom apps on league-issued hardware, not to every possible AI workflow team by team [10].
  • The specific use cases being restricted are live tactical decisions, not off-field scouting or general productivity [10].
  • The fact that a third of the league reportedly used the iPads this way suggests the policy could affect a meaningful share of teams’ workflows [10].
Story 031 source

Vertu is betting that executives will pay for an AI agent, not just a luxury phone

Vertu’s Alphafold is positioned as a luxury foldable for affluent buyers, especially chief executives, with prices starting at $6,880 and a pre-installed Hermes Agent that can analyze files, automate tasks across apps, remember conversations, and hand off requests to a human concierge [5]. In hands-on testing, the agent was good at some local-file and spreadsheet tasks and more willing than Samsung’s Gemini to automate multi-step workflows, but it also produced incomplete workflows, incorrect outputs, and inconsistent behavior [5].

Why it matters

This is a concrete sign that the premium-phone market is shifting from hardware status signaling to paid AI labor-saving features [5]. The review also highlights a broader product question for agentic AI: more autonomy can increase usefulness, but it can also raise reliability and user-control concerns [5].

Key insights

  • Vertu is explicitly targeting executives who want the phone to function as a working tool as much as a status object [5].
  • Hermes performed better than Gemini on some local-file and spreadsheet tasks in the reviewer’s testing [5].
  • The same autonomy that made Hermes feel more agent-like also led to errors and unfinished actions, showing the trade-off between independence and oversight [5].
  • The company appears to be actively updating the platform, so buyer experience may change quickly after launch [5].

Create your own daily briefing — start free. Keldura monitors the sources you choose and gives you a private, grounded daily digest with cited answers.

Create your own daily briefing — start free