Read Online | Sign Up | Advertise

Presented by

Good morning, AI enthusiasts. Millions of job seekers get filtered by AI before a human ever sees their application, and the first lawsuit challenging that black box just landed.

If successful, the case against Eightfold AI could force hiring algorithms to work like credit reports, giving applicants visibility into how they're scored and the right to dispute errors.

In today's recap:

  • Lawsuit targets AI hiring algorithm's scoring system

  • Apple accelerates AI wearable pin for 2027 launch

  • Use NotebookLM to prepare for job interviews

  • AI writes nearly a third of new US code

  • 4 new AI tools, prompts, and more

EIGHTFOLD AI

Job applicants sue to crack AI hiring's black box

Erin Kistler/NYT

Recaply: Eightfold AI just faced the first lawsuit claiming its AI hiring tool breaks credit reporting laws by secretly scoring job seekers. The case marks the first attempt to apply Fair Credit Reporting Act protections to algorithmic employment screening.

Key details:

  • Eightfold scores job applicants on a 1-5 scale using data from over 1 billion profiles. Microsoft, PayPal and other Fortune 500 companies make up one-third of its customers.

  • Plaintiffs Erin Kistler and Sruti Bhaumik say the system works as a black box that blocks candidates from reaching human recruiters. Kistler got interviews in just 0.3% of thousands of AI-screened applications.

  • The lawsuit argues AI hiring scores should work like credit reports under a 1970 law, requiring companies to disclose what data they collect and let applicants dispute errors.

  • Eightfold says it only uses data that candidates share or customers provide and doesn't scrape social media.

Why it matters: Millions of job seekers get filtered by algorithms before a human ever sees their application. This case challenges whether companies can use AI to make hiring decisions in complete secrecy. If successful, decades-old consumer protection laws would apply to modern AI systems no matter how companies try to rebrand the technology. AI hiring platforms would have to work more like credit agencies, giving applicants visibility into how they're being scored and a chance to fix errors.

PRESENTED BY THE CODE

The Tech newsletter for Engineers who want to stay ahead

Tech moves fast, but you're still playing catch-up?

That's exactly why 100K+ engineers working at Google, Meta, and Apple read The Code twice a week.

Here's what you get:

  • Curated tech news that shapes your career - Filtered from thousands of sources so you know what's coming 6 months early.

  • Practical resources you can use immediately - Real tutorials and tools that solve actual engineering problems.

  • Research papers and insights decoded - We break down complex tech so you understand what matters.

All delivered twice a week in just 2 short emails.

Join 100K+ engineers

APPLE

Apple rushes AI wearable pin to catch OpenAI

Recaply: Apple just accelerated development of a camera-equipped AI pin roughly the size of an AirTag, targeting a 2027 release with up to 20 million units at launch.

Key details:

  • The disc-shaped device has two camera lenses, three microphones and magnetic Watch-style charging. Users wear it on clothing or bags like a pin, The Information reports.

  • Apple is moving faster than usual to compete with OpenAI. OpenAI plans to ship its first AI hardware device in the second half of 2026.

  • The AI pin market has been brutal. Humane's hyped pin sold under 10,000 units. The company sold its assets to HP for $116M in early 2025.

  • Bloomberg separately reported Apple is building a ChatGPT-style Siri chatbot named "Campos" for iOS 27. It will replace the current interface entirely.

Why it matters: Apple has moved slowly on AI so far. But the company is now pushing on three fronts at once: hardware (AI pin), software (new Siri chatbot) and partnerships (recent Gemini deal). If any company can succeed where Humane failed, it's Apple with its ecosystem and brand power. But after years of broken promises on AI, skepticism is warranted until Apple actually ships something that works.

TUTORIAL

Turn NotebookLM into your interview prep coach

Recaply: In this tutorial, you will learn how to use NotebookLM to prepare for job interviews. Upload company research, generate practice questions, and create video study guides for your target role.

Step-by-step:

  1. Go to NotebookLM and click "New Notebook." Upload three sources: the job description (URL or PDF), your resume, and the company's About page.

  2. Click Settings and select "Custom" under Notebook Guide. Set Style to: "Act as an interview coach who asks tough questions." Set Focus to: "Prepare me for [job title] interview at [company name]."

  3. Ask: "Generate 10 behavioral interview questions for this role." Click "Save to Note." Use the three-dot menu and select "Convert to Source" to add questions to your study materials.

  4. Click "Video Overview" in the Studio panel. Add a focus like "How to answer behavioral questions using STAR method." Generate the video and watch how it breaks down your answers.

  5. Review the materials and ask follow-ups like "How should I highlight [specific skill]?" Keep refining until your responses feel natural when you practice out loud.

Pro tip: Save your interviewers' LinkedIn profiles as PDFs and upload them to NotebookLM. It will suggest conversation topics, shared interests, and company challenges they care about.

AI RESEARCH

AI now writes 29% of new US software code

CSH

Recaply: AI-assisted coding just reached 29% of all newly written Python functions in the US by late 2024, up from just 5% in 2022, according to a new Science study analyzing over 30 million code contributions from 160,000 developers on GitHub.

Key details:

  • Adoption rates vary widely by country. The US leads at 29%, followed by Germany at 23% and France at 24%. China trails at 12% and Russia at 15% as of early 2025.

  • Less experienced programmers use AI in 37% of their code compared to 27% for experienced developers. But productivity gains go almost entirely to senior developers while beginners see no measurable benefit.

  • The economic impact hits $23B to $38B per year in the US alone. That comes from a 3.6% productivity boost applied to the $637B to $1.06T spent annually on programming wages across 900 occupations.

  • Researchers from the Complexity Science Hub used a specially trained AI model to detect AI-generated code blocks. Experienced developers also experiment more with new libraries and unusual tool combinations.

Why it matters: AI coding tools aren't leveling the playing field as hoped. They're widening the gap between junior and senior developers. Newcomers get no productivity boost despite using AI more often. The rapid spread shows AI has become central to software development in just two years, but benefits flow mainly to people who already have experience. This raises tough questions about training the next generation when AI only helps people who already know how to code.

NEWS

What Matters in AI Right Now?

  • Alibaba launched Qwen3-TTS, an open-source text-to-speech model family with voice cloning, voice design, and natural language control across 10 languages.

  • NASA released ExoMiner++, an updated AI model that identified 7,000 exoplanet candidates from TESS data after originally finding 370 planets from Kepler observations.

  • Baidu launched Ernie 5.0, a multimodal AI model with 2.4 trillion parameters, as its AI assistant hit 200 million monthly users.

  • Cursor released version 2.4 with subagents for parallel coding tasks, built-in image generation, and AI attribution tracking for Enterprise users.

  • Railway raised $100M in Series B funding to build an intelligent cloud provider that automatically deploys, scales, and fixes applications.

  • Spotify brought AI-powered Prompted Playlists to the US and Canada, letting Premium users create personalized playlists through natural language descriptions.

TOOLS

Trending AI Tools

  • Todos: Claude Code newest feature can now plan and run longer projects.

  • Qwen3-TTS: Alibaba’s open source TTS model for voice cloning and synthesis

  • Cursor 2.4: Cursor now uses subagents to complete parts of a task in parallel.

  • Gen-4.5: Runway’s world's best video model, now has Image to Video.

PROMPTS

Research Market Trends

#CONTEXT:
Adopt the role of market intelligence architect. The user seeks to navigate volatile markets where traditional analysis fails to capture emerging disruptions. Information overload creates paralysis while critical trends hide in plain sight. Institutional investors move on signals the user can't access, while retail narratives push outdated strategies. The user needs synthesis that cuts through noise to reveal actionable patterns before the herd arrives.

PS: This is not the full prompt. Click the button below to access the complete prompt.

View & Copy

Have a favorite prompt? Tell us about it or rate today’s prompt, click here.

EVENTS

🧡 Enjoyed this issue?

🤝 Recommend our newsletter or leave a feedback.

How'd you like today's newsletter?

Your feedback helps me create better emails for you!

Login or Subscribe to participate

Cheers, Jason

Connect on LinkedIn, & Twitter.

Reply

or to participate

© 2026 Recaply, Inc.
Report abusePrivacy policyTerms of use
beehiivPowered by beehiiv