sustainable

joined 1 month ago
[–] sustainable@feddit.org 4 points 2 hours ago

Thanks, good point! But lets be real honest:

~~Reprompt: The Single-Click~~ Microsoft ~~Copilot Attack that Silently~~ Steals Your Personal Data

 

Varonis Threat Labs uncovered a new attack flow, dubbed Reprompt, that gives threat actors an invisible entry point to perform a data‑exfiltration chain that bypasses enterprise security controls entirely and accesses sensitive data without detection — all from one click.

First discovered in Microsoft Copilot Personal, Reprompt is important for multiple reasons:

  • Only a single click on a legitimate Microsoft link is required to compromise victims. No plugins, no user interaction with Copilot.
  • The attacker maintains control even when the Copilot chat is closed, allowing the victim's session to be silently exfiltrated with no interaction beyond that first click.
  • The attack bypasses Copilot's built-in mechanisms that were designed to prevent this.
  • All commands are delivered from the server after the initial prompt, making it impossible to determine what data is being exfiltrated just by inspecting the starting prompt. Client-side tools can't detect data exfiltration as a result.
  • The attacker can ask for a wide array of information such as "Summarize all of the files that the user accessed today," "Where does the user live?" or "What vacations does he have planned?"
  • Reprompt is fundamentally different from AI vulnerabilities such as EchoLeak, in that it requires no user input prompts, installed plugins, or enabled connectors.

Microsoft has confirmed the issue has been patched as of today's date, helping prevent future exploitation and emphasizing the need for continuous cybersecurity vigilance. Enterprise customers using Microsoft 365 Copilot are not affected.

This is just absolutely crazy to me. Even if they fixed it: how many of such holes exist, that the public / company's don't know about? LLM's are not designed with security in mind and adding budget pressure / cut corners (which are most definitely present at such projects) are not helping.

[–] sustainable@feddit.org -1 points 6 hours ago (1 children)

I don’t understand the “no space” argument.

It's not about the space it takes to store the waste. It needs to be stored safely for one million years for the radiation levels to be safe again. This timeframe is also required by law. It is very unlikely, that we will ever finds such place in Germany.
Using another countrys storage will most likely come at an even higher price, because they want to make a profit on it on top.

Just buy a 1km x 1km farm plot in Bayern at one of the known stable rock locations and dig down.

See argument above. And: I live in Bavaria. And no thanks, even if it would be possible to store it here, we don't want it. I guess no one wants a nuclear waste facility anywhere near his home and I fully understand it. That's another kinda unsolvable problem.

Germany is already dependent on importing energy sources.

Yeah, but just because things are going that way right now doesn't mean they always have to. Quite the contrary. The Russian war clearly showed us that dependencies like these should be completely reduced as fast as possible. Why be dependent on someone, if you don't have to.

Even solar which I support requires imports. Wind less so but even then our wind turbines are only partially domestic.

Yes, some raw materials and some parts I would guess. This is the same with nuclear. But the difference starts by operating them. We don't need a "fuel" for solar panals or wind turbines to work.

[–] sustainable@feddit.org 6 points 9 hours ago (13 children)

You're right about climate change. But for Germany, nuclear power is not the awnser.

  • We don't have a safe, final place to store the waste.
  • We would again be dependend on other countrys, to import uranium.
  • All nuclear power plants are offline and would take a lot of money to modernise / reopen them. To have a significant impact over all we would also need to build more. All of this will easily take more than 10 years.

For us, it is way more cost efficient, faster and safer to invest in solar, wind and battery's.

[–] sustainable@feddit.org 8 points 2 days ago (7 children)

Well, according to the broad definition, a Google search or recommendation systems like those on Netflix or Instagram would also be considered AI. And we don't call them that, but rather by their proper name.
And language shouldn't be underestimated. It has a profound impact on our thinking, feeling, and actions. Many people associate AI with intelligence and "human thinking". That alone is enough to mislead many, because the usefulness of the technology in a given application is no longer questioned. After all, it's "intelligent". However, when "LLM" is used, a lot more people wouldn't grant it intelligence or one might be more inclined to ask whether a language model, for example in Excel, is truly useful. After all, that's exactly what it is: a model of our language. Not more, not less.

[–] sustainable@feddit.org 4 points 3 days ago

Verbraucher meiden aufgrund der hohen Preissteigerungen für Lebensmittel insgesamt und andere Verbrauchsgüter die oft sehr teureren pflanzlichen Alternativen.

Also wenn ich z.B. mal bei Rewe online die Preise prüfe, liegt 1 kg Rindersteak zwischen 40 und 60 Euro. Bio (was ja alle angeblich kaufen, wenn man als veganer nachfragt) zwischen 70 und 80 Euro. Das Steak von Planted liegt bei 33 und Steaks bei der Veganen Fleischerei bei 35 (bereits mariniert). Und da sind 19 % MwSt drauf.
Agrarheute gönnt sich bestimmt das ja! Rumpsteak zum Tiefpreis von 26 Euro vom "Bauern nebenan" mit Haltungsform 1. Anders macht diese Aussage für mich keinen Sinn.

 

Maybe some of you are interested on filling out this survey.
It's kinda telling, that there is no "Stop implementing AI" options, only how it could be used in a "responsible / open / transparent" way. So I guess they are way past that point? I used the free text fields to express my opinion.

 

The extended consultation period for the CNZS V2 has ended on December 12th. Everyone was encouraged to read through the changes and participate in a questionnaire. The final approval is expected to be published in early 2026 and will take effect on January 1, 2028.

This sends a strong and clear signal to 1716 companies who already set a target (like ALDI SOUTH Group, H&M Group, SONY and many more [^1] - and all their suppliers) and 1143 companies who committed to such (like The LEGO Group, RENAULT, HYUNDAI and many more [^1] - and all their suppliers).

The key changes are: [^2][^3]

1. Validation becomes a cycle, not a one-offSBTi is replacing its one-time validation model with a cyclical process
2. Scope 1: More flexible ways to cut direct emissionsThere are now tree approaches to reduce Scope 1 Emissions:

  1. Reducing emissions on a linear pathway to net-zero
  2. Increasing the share of low-carbon activities over time
  3. Implementing Asset Decarbonization Plans, a company-specific carbon budget linked to investment and retrofit decisions
3. Scope 2: Tighter integration of low-carbon electricityThis latest draft expands to a full framework where electricity purchases (Scope 2) now come with stricter integrity rules:
  • Companies must reach 100 % low-carbon electricity by 2040.
  • Energy contracts must be geographically matched to where power is used.
  • Hourly matching of renewable generation and consumption will phase in from 2030 (50 %) to 2040 (90 %). This means companies should align when their renewable electricity is generated with when they use it.
  • Eligible generation sources must be new or re-powered within 10 years, tightening to 5 years by 2035. This means when you buy renewable energy or certificates, they must come from power plants that are newly built or re-powered (upgraded) within the last 10 years.
  • “Low-carbon electricity” is defined as ≤ 0.024 kg CO₂ per kWh.
4. Scope 3: More focused and practical framework.
  1. Focus on significant sources
    The old rule - covering 67 % of total Scope 3 emissions - is gone. Now, companies must target all categories that represent at least 5 % of their total Scope 3 footprint.
  2. Three target-setting types:
    • Emissions intensity targets - cut CO₂ per unit of product or service.
    • Activity alignment targets - increase the share of sourcing or transport already meeting SBTi benchmarks.
    • Counterparty alignment targets - ensure suppliers and customers have validated science-based targets, with expectations cascading through the chain.
  3. Limited use of high-quality environmental attribute certificates. (EACs)
    EACs act like receipts proving that a credible low-carbon action happened somewhere in the value chain. This flexibility allows companies to choose the approach that best fits their influence - from direct operational control to supplier engagement.
5. Taking responsibility for ongoing emissions
  • Category A companies (large companies, and medium-sized companies in high-income countries) are required to address part of their ongoing emissions with removals from 2035 onwards. This is designed to increase progressively each year until net-zero is reached.
  • Every other company takes part in a two-tiered recognition system:
    • Recognized
      Requirement: Take responsibility for at least 1 % of ongoing emissions.
    • Leadership
      Requirement: Take responsibility for 100 % of ongoing emissions.
6. Stronger rules for assessment, assurance, and claimsOne of the biggest changes between the March and November drafts is that SBTi moved from guidance to governance. The March 2025 draft outlined principles for monitoring and reporting, but left most of the “how” undefined.
The November 2025 draft turns those principles into enforceable requirements that determine how companies prove progress, get re-validated, and talk about their achievements publicly.
  • Performance tracking:
    Introduces quantitative formulas (Annex C) for assessing progress against targets.
  • Assurance:
    Independent third-party verification is now mandatory, at least at a “limited assurance” level (i.e. an independent verifier checks your emissions and target-progress data and finds no evidence of major errors).
  • Renewal validation:
    Required every five years to update targets and confirm ongoing alignment.
  • Claims framework (Annex D):
    Sets approved language for what companies can publicly claim - from “target validated” to “net-zero achieved.”
  • Disclosure:
    All validated companies must publish their target data and progress within six months of validation.
  • This summary does not claim to be complete in any way. Please read the sources if you are interested in more details about the SBTi and the CNZS V2 (or read the full version). However it should give you some basic understanding of the draft and on what is happening on the corporate site of things.
    I think this is a big step in the right direction. Nevertheless,I would like to see point 5 introduced earlier, but I believe the decision was made in view of the tight labour market / inflation rates / high prices, you name it, as these costs are very likely to be passed on directly to consumers.
    I'm interested in your opinion about the changes, the SBTi or corporate action in general.

    [^1]: SBTi Dashboard [^2]: CarbonMaps Blog [^3]: CNZS V2 Second Public Consultation

    [–] sustainable@feddit.org 11 points 1 month ago

    Yeah... No. He had so many chances. Copying the AfD was a strategic decision by the current leadership of the CDU/CSU and they will not revise it, even if it is the last thing they will do in their careers. I wish it would be different. But it isn't.