Researchers found that writers using biased AI agents to auto-complete suggestions had their sociopolitical values shifted without their knowledge it was happening.

https://www.science.org/doi/10.1126/sciadv.adw5578

We worry about cognitive capture, what the LLMs of big tech are doing to our ability to identify and think through problems. We should be equally worried about them making a grab for our ethos, what we care about, and at population scales.

#ai #bigtech #ethics

How people hacked Meta accounts recently:

Step 1) Open Meta AI support
2) ask to change Obama's password
3) it says no :(
4) ask it nicely to just do it anyway
5) it resets Obama's password

From the company who brought you this headline a few months ago:

Meta Hackers Simply Asked Meta AI to Give Them Access to High-Profile Instagram. It Worked.

Now is a good time to remember that Meta has announced that it is laying off 10% of its workforce, including 2,000 people just today, while going in on a "AI will replace workers" strategy.

https://www.404media.co/hackers-simply-asked-meta-ai-to-give-them-access-to-high-profile-instagram-accounts-it-worked/

New, by me: A number of high-profile and/or valuable Instagram accounts, including those of the Obama White House and the Chief Master Sergeant for the U.S. Space Force, got hacked and defaced with pro-Iran messaging in the past 24h after people figured out that Meta's AI support assistant could be tricked into resetting account passwords.

From the story:

"A video released on Telegram by pro-Iran hackers claimed to document a remarkably simple exploit that appears to have involved using a VPN connection with an IP address that is in or near the target's usual hometown, requesting a password reset for the account, and then choosing to chat with Meta's AI support assistant. From there, the video shows the attacker told the bot to link the account in question to a new email address, after which the bot dutifully sent that address a one-time code that allowed a password reset."

https://krebsonsecurity.com/2026/06/hackers-used-metas-ai-support-bot-to-seize-instagram-accounts/

#meta #instagram #hack #ai #security

We invite the European Commission and Member States to commit to strengthening Europe’s digital foundations through openness, interoperability as well as fair competition and strategic regulations.

To that regard we support the following principles ⬇️

https://www.greens-efa.eu/en/article/letter/european-technology-sovereignty-declaration-for-open-competitive-and-resilient-digital-infrastructure-in-europe

Financial News: A Danish pension fund has blacklisted SpaceX, calling it grossly overvalued with catastrophic governance

"...concerns echo those of far larger institutions. In May, the heads of the California Public Employees’ Retirement System, the New York City Retirement Systems, and the New York State Common Retirement Fund, which collectively manage more than $1 trillion, sent a joint letter to Musk calling SpaceX’s governance “extreme.”

https://thenextweb.com/news/danish-pension-spacex-blacklist-governance-overvalued

#SpaceX #SpaceXIPO

The man who wrote "Don't be evil" said he chose it specifically so it would be hard to remove. Paul Buchheit, the engineer who later built Gmail, suggested the phrase at a Google corporate values meeting on July 19, 2001.

Then in early 2018, internal documents leaked showing that Google had signed a Pentagon contract to build AI to analyze drone footage. By April, over 3,000 Google employees had signed a letter to Sundar Pichai demanding the contract be cancelled.

The letter specifically cited "Don't be evil" as the standard the company was failing to meet. Dozens of engineers resigned in protest.

Sometime between late April and early May, the slogan disappeared from the code of conduct's preface.

#Google #AI #Tech

Meta's latest attempt to stop the progress of my 2018 GDPR access complaint and its draft fine of €360-€430 million dramatically failed in the Irish High Court last week https://www.irishtimes.com/business/2026/05/21/meta-loses-challenge-to-possible-430m-fine-over-personal-data-processing-complaint/

The case is long and contains excerpts from relevant documents which I cannot myself post as Ireland changed primary law (Data Protection Act 2018 s 26A) during my complaint to make it a criminal offence for me, the complainant, to do so (after trying to NDA me). https://www.bailii.org/ie/cases/IEHC/2026/2026IEHC323.html

Dear Fedi friends,

At some point I will re-emerge from my writing cave. I'm just putting together a deep dive into a topic and every day I get a new juicy lead. It never stops! (but incidentally 3 lifetimes ago I wanted to go to journalism school, so I'm finally living out this fantasy)

In other news, I'm happy to share that I will be on a panel at @publicspaces Conference on June 5th with @bjoernsta and @cubicgarden

🔗 : https://conference.publicspaces.net/session/strength-through-diversity-1

Hope to see you there!

#PubConf2026