💼If you use AI and are a woman, your code is perceived worse💼

AI at work was supposed to boost productivity, but new research shows it’s also boosting something else: bias.

A study of 28,698 software engineers at a major tech company found AI adoption stubbornly low (just 41% overall, dropping to 31% among women and 39% among engineers over 40). Why? Not because of access or training gaps—but because of what researchers call the “competence penalty.”😯

• In experiments with 1,026 engineers, identical code was rated as worse when reviewers thought AI helped write it.

• Engineers were judged 9% less competent on average if they used AI—even though code quality was identical.

• For women, the penalty jumped to 13%, nearly double the 6% for men.

• The harshest critics? Male non-adopters, who penalized female AI users 26% more harshly than men.

This creates a chilling effect: the very groups who could benefit most from AI (women, older workers) use it least, fearing career damage.😤

This “competence penalty” is more than hurt feelings—it’s lost profit, hidden bias, and shadow AI use:

• Companies miss out on billions in ROI when licenses sit idle.

• Employees use unsanctioned AI tools in secret, risking security and compliance.

• Bias deepens—AI becomes another way to question the skills of underrepresented groups.

It’s the opposite of the promise that AI would “level the playing field.”😣

The research offers a roadmap:

1. Map the hotspots: Find where penalties fall hardest (often in teams with skewed demographics and power imbalances).

2. Flip the influencers: Non-adopters impose the harshest penalties. Visible role models—especially senior women and underrepresented leaders—must use AI openly to normalize it.

3. Redesign evaluation: Stop tagging AI-assisted work in reviews. Judge outputs, not whether AI was in the room. Some companies (👀 Microsoft, Shopify) are even rewarding AI usage in performance reviews.

Read more here in the HBR

🛍️Off target🛍️

Target just learned the hard way what happens when you treat diversity as a nice to have.👀

In January, after Trump’s executive order railing against DEI, Target quietly announced it was rolling back its equity initiatives. Consumers saw it as abandonment. Employees saw betrayal. Investors? They saw red.📉

The fallout was brutal:

• $12 billion wiped off investor value in weeks.

• Stock price down 24% almost overnight.

• Boycotts disguised as “fasts” spread nationally, with over 200,000 participants.

• Trust tanked.

• And this week: CEO Brian Cornell is out.

The irony? DEI had built Target’s boom. After George Floyd’s murder in its home state, Target pledged billions in spend with Black-owned businesses, invested in Black founders, and visibly tied its growth to equity. Sales surged, market cap hit $129B, and brand trust soared. Fast forward to 2025: ditching DEI has halved Target’s value (now $45B) and shredded the goodwill it took decades to build💸

This is more than a retail story. It’s proof that DEI isn’t “woke window dressing”—it’s strategic infrastructure for trust, growth, and resilience. Target’s about-face shows what happens when companies treat values as negotiable: customers leave, employees disengage, and markets punish.😐

Read more: Forbes coverage

🖥️Tech Tariff wars🖥️

President Trump has opened a new front in the tech trade wars—this time targeting Europe’s digital taxes and the Digital

Services Act (DSA).🧑🏾‍⚖️

• Trump threatened “substantial additional tariffs” on exports from any country that taxes U.S. tech giants like Google, Meta, Apple, or Amazon.

• The administration is also considering sanctions on EU officials implementing the DSA, the bloc’s sweeping online safety law. That could mean visa bans—an unprecedented step against allies.

• Washington’s line: digital taxes and the DSA are “discriminatory” and designed to censor Americans, particularly conservative voices online.

• Europe’s line: the DSA is about removing illegal content (hate speech, child exploitation, disinformation) while safeguarding free expression.

This is not just regulatory squabbling—it’s a clash of industrial strategy vs. sovereignty:🇺🇸🇪🇺

• For the U.S., digital services are one of its most profitable exports. Taxes and rules abroad are framed as attacks on “American innovation.”

• For the EU, the DSA is a cornerstone of its tech sovereignty project—forcing Big Tech to play by European rules on safety, competition, and responsibility.

• If tariffs and sanctions escalate, we could see the first real “tech trade war”—hitting everything from consumer goods to semiconductor access.

The bigger picture🔍

• For tech companies, this means living in two regulatory universes: Europe’s stricter guardrails vs. Washington’s hands-off, protect-the-home-team stance.

• For startups and investors, uncertainty around tariffs and compliance could ripple far beyond Big Tech—affecting valuations, market access, and cross-border deals.

• For policymakers, it’s a stress test of whether “global tech rules” can exist at all, or if we’re sliding into fragmented digital blocs.

📉 So what?

This is more than Trump’s usual tariff theatrics—it signals a world where tech regulation is weaponised in trade policy. Europe wants to shape global norms; the U.S. wants to shield its giants. The danger is that smaller players—startups, emerging markets, innovators— and maybe even countries cough the UK get crushed in the middle.

📊A slice of the pie?📊

File this one under “late-stage industrial strategy.” The US government just bought a 9.9% stake in Intel, dropping $8.9 billion into common stock—on top of the $2.2 billion in CHIPS Act grants Intel already banked. Total tab? $11.1 billion.💰

Intel framed it as a confidence boost in its turnaround, but let’s be real: this is also about Washington hedging against foreign chip dominance (hello, Taiwan’s TSMC) and protecting supply chains in an AI-obsessed world where chips are gold dust.

President Trump (back in dealmaker mode) announced the agreement after meeting Intel’s CEO, Lip-Bu Tan—whom he’d recently tried to shove out over China ties. Now, instead of firing him, Trump’s praising his “amazing story” while claiming the US “picked up $10 billion.” Classic.😏

The fine print:

• The US gets nearly 433M Intel shares at a discount.

• No board seats, no governance rights—it’s passive ownership (at least on paper).

• Intel shares popped 6% after the news, though slipped in after-hours trading.

This isn’t just a cash infusion—it’s the government owning a chunk of Big Tech. Intel is one of the last US companies still designing and manufacturing advanced chips domestically. With AI, defense, and every iPhone in your pocket depending on silicon, Washington’s message is clear: “Too big to outsource, too strategic to fail.”😯

It also signals a bigger trend: the US is taking golden shares and equity stakes in strategic industries—from steel to rare earths—nudging capitalism toward something more… state-curated. Not everyone is a fan with critics noting it may stifle company innovation and is more akin to state planned enterprises more common in say the USSR😦

For Asia and Europe watching semiconductor policy, this is the playbook: throw money and ownership at your national champions, or get left behind.

You can see the coverage of it here

🧠So what?

Government-as-shareholder raises new questions: how will public interest shape corporate strategy when taxpayer cash is on the line? And will this model spread to AI firms or cloud infrastructure next? Intel might just be the warm-up act for a new era of state-capital entanglement.

💸Grok-tacular leak💸

Nearly 300,000 user conversations—meal plans, medical questions, password requests, even instructions for cooking up Class A drugs—were sitting pretty on Google search. Why? Because Grok’s “share chat” button didn’t just share with your mate, it also handed the transcript to the entire internet as the URL is public and indexable.💻

And it’s not just Grok. OpenAI had a similar issue earlier this year. Turns out AI chatbots are great at generating text… and terrible at keeping it private.😅

• Personal data risk: Even without names, shared prompts often reveal health details, business info, or location. Once indexed, that data is forever searchable.

• Trust gap: Users aren’t being told their data could be public—undermining confidence in AI tools just as they go mainstream.

• Regulatory spotlight: Expect watchdogs (and lawyers) to notice. If your bot leaks, it’s not just a bug—it’s a compliance nightmare.

Oxford’s Prof. Luc Rocher called chatbots a “privacy disaster in progress.” Hard to argue when one slip of a button can turn private queries into Googleable headlines.😐

🧠So what?

If AI chats can leak this easily, it’s not just a Grok problem—it’s a preview of what happens when tech companies rush. Today it’s your meal plan on Google; tomorrow it could be your business strategy, health data, or student records. The bigger picture? AI can’t scale if trust doesn’t. If companies can’t guarantee privacy, adoption stalls, regulation spikes, and the AI “revolution” risks collapsing under its own leaks.

📈 The tools behind the tech📉

📦Product📦

• Code review can be better
• Good system design

📏Design📏 

• Interactive guide to SVG paths 

👩🏿‍💻Code👩🏿‍💻

• Vibe coding tips and tricks

🏢The business behind the tech🏢

•  Engineering blogs from big companies
• A case study on bad hiring

AI for everyone

Boost your career with free AI training – Apply now!
Colorintech is offering free AI courses to Everyone as part of
the Google.org AI Opportunity Fund: Europe. 

This course will equip you with the necessary knowledge and practical experience to succeed in today’s AI driven job market. 

If you’re looking to upskill in AI, boost your employability, and
open doors to exciting career opportunities, this is your chance! 

Applications are open now — visit our website to apply 

BTF

Munya Chawawa is speaking at this year's BTF

Get your tickets now

Here is more info: joinbtf.com

Amplify with Bloomberg

Discover Innovative and Exciting Insights from Bloomberg’s Data and Engineering Teams

On Sep 10th, Bloomberg is joining forces with Colorintech for another installment Amplify, an event dedicated to amplifying diverse voices in the world of tech!

This evening event will feature a series of dynamic lightning talks and a special panel discussion covering a range of cutting-edge topics from leading voices at Bloomberg.

Following the talks and panel, you’ll also have the opportunity to connect with speakers, leaders, and other members of the Colorintech community during our networking session, whilst enjoying delicious food and drinks.

This event is ideal for those with experience in:

Data, including; Data Engineer, Data Modelers and Data Quality

Software Engineering

AI Engineering.

This is a great opportunity for anyone who wants to learn, be inspired, and build connections with Bloomberg Professionals who are shaping the future of technology!

Check out the key details below: 

Date: Wednesday 10th Sep 2025

Time: 17:30 - 20:30 UK Time

Where: Central London

As this event has limited spaces, we'll need you to register your interest to attend using our application form below which closes on the 26th August: 

https://bloomberg.avature.net/su/1c28073f3478fe19

After applying, the Bloomberg team will reach out to you and provide you with next steps if they have invited you along.

Cornerstone's VC fellowship

Step into the world of venture capital with Cornerstone NextGen, a 9-month fellowship that offers a firsthand look at VC through the lens of an emerging fund.

This programme is for current students or those who have graduated from a UK university in the past 12 months.

Link here

😃What we are consuming😃

📱Google's newest devices