Toddzilla

For those that don't subscribe, I encourage you to do my so. Valuable stuff that is worth every penny.

Announcing the "AI Agent Standards Initiative" for Interoperable and Secure Innovation The Initiative will ensure that the next generation of AI is widely adopted with confidence, can function securely on behalf of its users, and can interoperate smoothly across the digital ecosystem.

NIST is stepping into standards for AI agents. We have MCP (and OAuth), A2A, and ancillary things like mutual TLS. But there is a need for refinement (lots of questions about OAuth meeting requirements). #cybersecurity #identity www.nist.gov/news-events/...

This shows the sort of data that Proton Mail, which prides itself both on end-to-end encryption and that it's only governed by Swiss privacy law, can provide to third parties. In this case, the FBI. @evystadium.bsky.social has more.

Scoop by @josephcox.bsky.social: www.404media.co/proton-mail-...

Silicon Valley isn't monolithic. The plutocrat/billionaire/Epstein class (Zuckerberg/Andreesson/etc) have a different perspective than the AI foundation model companies, which is different than lots of other Silicon Valley tech sectors, etc.

Netflix will pocket a hefty breakup fee, no? They may not have won the bidding, but they don't walk away with nothing.

AI Recommendation Poisoning: When Optimization Becomes Manipulation In this episode of the Microsoft Threat Intelligence Podcast, ⁠Sherrod DeGrippo⁠ speaks with Microsoft security and AI researchers Giorgio Severi and Noam Kochavi about a newly observed trend in AI abuse: ⁠recommendation poisoning through memory manipulation⁠. While looking into prompt injection and reprompt-style behaviors, the team uncovered something quieter but potentially more persistent—websites embedding hidden instructions inside Summarize with AI links that attempt to influence what an AI assistant remembers and recommends over time. Rather than focusing on immediate exploitation, this technique aims to shape long-term behavior inside AI systems. Giorgio and Noam explain how it works, why it’s spreading across industries, where legitimate marketing tactics can blur into security risk, and what defenders and users should understand about managing AI memory in an increasingly agent-driven environment.

AI systems are increasingly becoming decision support systems, and threat intelligence shows that their memory could be deliberately influenced. msft.it/63329Qwrsl

Meta Director of AI Safety Allows AI Agent to Accidentally Delete Her Inbox Meta Superintelligence Labs’ director of alignment called it a “rookie mistake.”

AI agents are a huge expansion of the enterprise cybersecurity attack surface with a lot of room for mischief and errors. The latest example of an error: www.404media.co/meta-directo...

And good luck to everyone in finding the budget to overcome the tech debt around rewriting and then QAing the code.

An AI coding bot took down Amazon Web Services Blames "user error, not AI error" for incident in December involving its Kiro tool.

Identity & AWS outage:
"Amazon said that by default its Kiro tool “requests authorisation before taking any action” but said the engineer involved in the December incident had “broader permissions than expected—a user access control issue, not an AI autonomy issue.”"
arstechnica.com/ai/2026/02/a...

“I feel like, increasingly, it's impossible to talk to a single human [at Pinterest],” Pinterest user Tiana Oreglia said. “Along with being filled with AI images that have been completely ruining the platform, Pinterest has implemented terrible AI moderation” www.404media.co/pinterest-is...

2026 Unit 42 Global Incident Response Report Read the 2026 Unit 42 Global Incident Response report to discover attacker tactics and get real-world insights and expert recommendations to safeguard your business.

There is job security that comes with working in identity security/IAM. From Palo Alto Networks Unit 42: "Identity has become the most reliable path to attacker success. Identity weaknesses played a material role in almost 90% of Unit 42 investigations." www.paloaltonetworks.com/resources/re...

CFP | NA 2026 | fwd:cloudsec fwd:cloudsec is a non-profit conference on cloud security. At this conference you can expect discussions about all the major cloud platforms, both attack and defense research, limitations of security...

Time to submit for cloudsec if you haven't already: fwdcloudsec.org/conference/n...

2 to 3 Cups of Coffee a Day May Reduce Dementia Risk. But Not if It’s Decaf.

I plan to increase my coffee intake, and forget that decaf stuff. www.nytimes.com/2026/02/09/h...

Analysis of active exploitation of SolarWinds Web Help Desk | Microsoft Security Blog We are seeing exploitation of SolarWinds Web Help Desk via CVE‑2025‑40551 and CVE‑2025‑40536 that can lead to domain compromise; here is how to patch, hunt, and mitigate now.

Microsoft Defender Research has published details, mitigation, detection, and hunting guidance on the observed exploitation of internet‑exposed SolarWinds Web Help Desk (WHD) systems: msft.it/63327QPD9N

Como dicen en México, no mames! Spanish has a long run ahead in influencing American culture. Nice headline to get attention, though.

Trump’s Big Loser Energy, and Other Tales From the Annals of Political Messaging A few days ago Donald Trump said he’s deciding to “nationalize” American...

Play your game, not the other guy's game. Outrage ain't the way - I'd go for contempt, scorn, and laughter. The guy is a losing, lying weasel. talkingpointsmemo.com/edblog/trump...

On iPhone: five rapid taps in succession on the lock button shuts off all biometrics.

So sorry to see this happen @joemenn.bsky.social - you are a superstar. One door closes and others open. I look forward to reading your news of the next adventure.

FBI Couldn’t Get into WaPo Reporter’s iPhone Because It Had Lockdown Mode Enabled Lockdown Mode is a sometimes overlooked feature of Apple devices that broadly make them harder to hack. A court record indicates the feature might be effective at stopping third parties unlocking some...

New from 404 Media: the FBI has been unable to get into the iPhone of raided Washington Post journalist because the phone had Lockdown Mode enabled. Apple markets Lockdown Mode mostly to stop spyware like NSO. Here, a real world example of it stopping access too www.404media.co/fbi-couldnt-...

Lockdown Mode - use it or risk losing control of your data if someone is able to physically take your IOS (iPhone) device.

FBI Couldn’t Get into WaPo Reporter’s iPhone Because It Had Lockdown Mode Enabled Lockdown Mode is a sometimes overlooked feature of Apple devices that broadly make them harder to hack. A court record indicates the feature might be effective at stopping third parties unlocking some...

Exhibit 1 on why IOS Lockdown Mode is useful: www.404media.co/fbi-couldnt-...

Pondering on this, LinkedIn is like print media of old - you have tabloids (mediocrity) and reputable new sources (NYT, WSJ, FT,...). It depends what you read... or in case of LinkedIn, who you follow or are connected with. Of course, we also have the LI algo injecting rubbish ads into our feeds...

Where is a reputable gofundme to help defray legal costs?

Expect these strategies to proliferate to other professional leagues if they succeed consistently. Fortunately my Sunday league of soccer duffers has to worry about showing up for work on Monday uninjured - nobody wants to get hurt trying to crowd the keeper.

Some things you should know about Fulton County, Georgia The county has already seen numerous claims about election fraud — all debunked. There's no reason to think that Kash Patel's FBI will find something new.

I felt obliged to write up an overview of all of the Fulton County fraud claims that have already been made — and dismissed. Let me know if I forgot any. www.pbump.net/o/some-thing...

People, think carefully before putting your sensitive personal information into random apps. They frequently have vulnerabilities and can leak your information. See below for the latest example courtesy of @404media.co .

Umm, why isn't the local police department ticketing the vehicles for violating state law by not having license plates?

Bitte, mach es!

On Tilt, by Jasper Craven America’s new gambling epidemic

"Nearly half of all American men aged 18 to 49 maintain an online sports-betting account...Emerging research suggests that the spread of sports gambling portends a huge increase in gambling addiction, which has the highest rate of suicide of any addictive behavior."

harpers.org/archive/2026...

Security Engineer 2 - Red Team | Microsoft Careers Discover and exploit vulnerabilities end-to-end in order to assess the security of systems and services Advocate for security change through building partnerships and clearly communicating impact of r...

Want to red team azure?