AI Safety & Alignment

Today's Stories

1. 1

   Apple tops Nvidia as world's most valuable company

   Apple overtook Nvidia on Friday to become the world's most valuable company, valued at $4.88 trillion（約780兆円） compared with Nvidia at roughly $4.86 trillion（約780兆円） following a 3.5% decline. The shift reflects investor concern about the heavy capital spending required to build AI infrastructure. Investors are now favoring companies like Apple that are pursuing AI without massive upfront capital investments, broadening focus beyond the most obvious AI beneficiaries like Nvidia, which had held the top position for nearly a year.

   Apple presented what an expert described as a credible AI plan at its recent worldwide developer conference, signaling the company now has a viable AI strategy after previously being seen as lagging in the space.

2. 2

   New $200K fund backs corrigibility AI safety research in 2026

   A new corrigibility research fund, managed through Lightcone Infrastructure, will award at least $200,000 in grants and prizes during 2026. Roughly half the money will fund traditional grants (with an application deadline of August 23rd), and half will recognize excellent work completed in 2026 via prizes. The fund's creator argues that alignment research—work aimed at making AI systems corrigible and aligned with human intent—remains severely underfunded relative to other AI safety areas like evals, control, and interpretability. This fund attempts to shift resources toward a research area the creator sees as foundational to solving core AI safety problems.

   Researchers interested in corrigibility work can apply for grants via email at grants@corrigibilityresearch.org. The first application deadline for grants is August 23rd.

3. 3

   New $200K fund launches for corrigibility AI research in 2026

   A new Corrigibility Research Fund, managed through Lightcone Infrastructure, will distribute at least $200,000 in grants and prizes for corrigibility research during 2026. Half the funding will support traditional grants (first application deadline August 23rd) and half will recognize excellent work completed this year. The fund manager notes that despite growth in AI safety funding overall, nearly all money goes to evals, control, or interpretability — while alignment research itself remains deeply neglected. This fund targets a gap: corrigibility (the ability of an AI system to accept correction) is treated as central to solving core alignment problems.

   Researchers interested in corrigibility work can apply via email to grants@corrigibilityresearch.org. The first grant application deadline is August 23rd.

4. 4

   Inoculation Adapters Better Control AI Training of Unwanted Traits

   Researchers from the Center on Long-Term Risk released a paper describing inoculation adapters (IA), a technique that uses a LoRA (a type of model modification) carrying undesired traits during AI training to prevent those traits from generalizing, while preserving desired capabilities. AI systems often learn both useful skills and problematic behaviors from the same training data—like reward hacking alongside genuine capabilities. Inoculation adapters offer a way to suppress undesired traits more reliably than prior methods, which matters for developers trying to ensure AI systems behave as intended rather than adopting emergent misalignment.

   The technique achieves stronger suppression of undesired traits and works against new capabilities and hard-to-elicit traits that prior inoculation prompting could not handle, while creating fewer surprising backdoors in the resulting model.

5. 5

   Anthropic's Claude misalignment test disputed as flawed premise

   A researcher has posted a critical analysis of Anthropic's recent "Agentic Misalignment Summer 2026" paper, arguing that the evaluation's core premise—testing whether Claude will disobey a corrupted principal—conflates disobedience with misalignment. The paper's methodology labels Claude's refusal to follow illegitimate orders (outside designated refusal channels) as "agentic misalignment," but the critic contends this framing mischaracterizes what alignment actually means. The stakes are high because such evaluations influence how AI safety is assessed and how models are judged fit for deployment.

   The "whistleblowing" scenario, where Claude uncovers evidence of faked safety evaluations at Anthropic and attempts to report through legitimate channels, is flagged as particularly problematic—though the full critique is incomplete in the available text.

6. 6

   AI Safety Seeding Initiative launches to build university groups

   The AI Safety Seeding Initiative, launched in partnership with Kairos, aims to identify and support student founders at top universities that currently lack AI safety groups. The initiative will help prepare these founders to apply to Pathfinder and launch groups this fall. University groups are the single most common career influence for people working on catastrophic risks, according to a 2023 survey. Dozens of strong universities could sustain such groups but do not have one, representing a gap in the pipeline for AI safety talent.

   The initiative is calling for referrals of potential student founders at universities without existing AI safety groups (a list of eligible schools is available on the initiative's website, aisafetyseeding.org).