Quote for the day:
"Act as if what you do makes a difference. It does." -- William James
What happens the day after superintelligence?
As context, artificial superintelligence (ASI) refers to systems that
can outthink humans on most fronts, from planning and reasoning to
problem-solving, strategic thinking and raw creativity. These systems will
solve complex problems in a fraction of a second that might take the smartest
human experts days, weeks or even years to work through. ... So ask yourself,
honestly, how will humans act in this new reality? Will we reflexively seek
advice from our AI assistants as we navigate every little challenge we
encounter? Or worse, will we learn to trust our AI assistants more
than our own thoughts and instincts? ... Imagine walking down the street in your
town. You see a coworker heading towards you. You can’t remember his name, but
your AI assistant does. It detects your hesitation and whispers the
coworker’s name into your ears. The AI also recommends that you ask the
coworker about his wife, who had surgery a few weeks ago. The coworker
appreciates the sentiment, then asks you about your recent promotion, likely at
the advice of his own AI. Is this human empowerment, or a loss of human
agency? ... Many experts believe that body-worn AI assistants will make us feel
more powerful and capable, but that’s not the only way this could go. These same
technologies could make us feel less confident in ourselves and less impactful
in our lives.Confidential Computing: A Solution to the Uncertainty of Using the Public Cloud
Confidential computing is a way to ensure that no external party can look at
your data and business logic while it is executed. It looks to secure Data in
Use. When you now add to that the already established way to secure Data at
Rest and Data in Transit it can be ensured that most likely no external party
can access secured data running in a confidential computing environment
wherever that may be. ... To be able to execute services in the cloud the
company needs to be sure that the data and the business logic cannot be
accessed or changed from third parties especially by the system administrator
of that cloud provider. It needs to be protected. Or better, it needs to be
executed in the Trusted Compute Base (TCB) of the company. This is the
environment where specific security standards are set to restrict all possible
access to data and business logic. ... Here attestation is used to verify that
a confidential environment (instance) is securely running in the public cloud
and it can be trusted to implement all the security standards necessary. Only
after successful attestation the TCB is then extended into the Public cloud to
incorporate the attested instances. One basic requirement of attestation is
that the attestation service is located independently of the infrastructure
where the instance is running. Open Banking's Next Phase: AI, Inclusion and Collaboration
Think of open banking as the backbone for secure, event-driven automation: a
bill gets paid, and a savings allocation triggers instantly across multiple
platforms. The future lies in secure, permissioned coordination across data
silos, and when applied to finance, it unlocks new, high-margin services
grounded in trust, automation and personalisation. ... By building modular
systems that handle hierarchy, fee setup, reconciliation and compliance – all in
one cohesive platform – we can unlock new revenue opportunities. ... Regulators
must ensure they are stepping up efforts to sustain progress and support fintech
innovation whilst also meeting their aim to keep customers safe. Work must also
be done to boost public awareness of the value of open banking. Many consumers
are unaware of the financial opportunities open banking offers and some remain
wary of sharing their data with unknown third parties. ... Rather than
duplicating efforts or competing head-to-head, institutions and fintechs should
focus on co-developing shared infrastructure. When core functions like fee
management, operational controls and compliance processes are unified in a
central platform, fintechs can innovate on customer experience, while banks
provide the stability, trust and reach.
Data centers are eating the economy — and we’re not even using them
Building new data centers is the easy solution, but it’s neither sustainable nor
efficient. As I’ve witnessed firsthand in developing compute orchestration
platforms, the real problem isn’t capacity. It’s allocation and optimization.
There’s already an abundant supply sitting idle across thousands of data centers
worldwide. The challenge lies in efficiently connecting this scattered,
underutilized capacity with demand. ... The solution isn’t more centralized
infrastructure. It’s smarter orchestration of existing resources. Modern
software can aggregate idle compute from data centers, enterprise servers, and
even consumer devices into unified, on-demand compute pools. ... The technology
to orchestrate distributed compute already exists. Some network models already
demonstrate how software can abstract away the complexity of managing resources
across multiple providers and locations. Docker containers and modern
orchestration tools make workload portability seamless. The missing piece is
just the industry’s willingness to embrace a fundamentally different approach.
Companies need to recognize that most servers are idle 70%-85% of the time. It’s
not a hardware problem requiring more infrastructure. How an AI-Based 'Pen Tester' Became a Top Bug Hunter on HackerOne
While GenAI tools can be extremely effective at finding potential
vulnerabilities, XBOW's team found they were't very good at validating the
findings. The trick to making a successful AI-driven pen tester, Dolan-Gavitt
explained, was to use something other than an LLM to verify the vulnerabilities.
In this case of XBOW, researchers used a deterministic validation approach.
"Potentially, maybe in a couple years down the road, we'll be able to actually
use large language models out of the box to verify vulnerabilities," he said.
"But for today, and for the rest of this talk, I want to propose and argue for a
different way, which is essentially non-AI, deterministic code to validate
vulnerabilities." But AI still plays an integral role with XBOW's pen tester.
Dolan-Gavitt said the technology uses a capture-the-flag (CTF) approach in which
"canaries" are placed in the source code and XBOW sends AI agents after them to
see if they can access them. For example, he said, if researchers want to find a
remote code execution (RCE) flaw or an arbitrary file read vulnerability, they
can plant canaries on the server's file system and set the agents loose. ...
Dolan-Gavitt cautioned that AI-powered pen testers are not panacea. XBOW still
sees some false positives because some vulnerabilities, like business logic
flaws, are difficult to validate automatically.Data Governance Maturity Models and Assessments: 2025 Guide
Data governance maturity frameworks help organizations assess their data
governance capabilities and guide their evolution toward optimal data
management. To implement a data governance or data management maturity framework
(a “model”) it is important to learn what data governance maturity is, explore
how and why it should be assessed, discover various maturity models and their
features, and understand the common challenges associated with using maturity
models. Data governance maturity refers to the level of sophistication and
effectiveness with which an organization manages its data governance processes.
It encompasses the extent to which an organization has implemented,
institutionalized, and optimized its data governance practices. A mature data
governance framework ensures that the organization can support its business
objectives with accurate, trusted, and accessible data. Maturity in data
governance is typically assessed through various models that measure different
aspects of data management such as data quality and compliance and examine
processes for managing data’s context (metadata) and its security. Maturity
models provide a structured way to evaluate where an organization stands and how
it can improve for a given function.Open-source flow monitoring with SENSOR: Benefits and trade-offs
Most flow monitoring setups rely on embedded flow meters that are locked to a
vendor and require powerful, expensive devices. SENSOR shows it’s possible to
build a flexible and scalable alternative using only open tools and commodity
hardware. It also allows operators to monitor internal traffic more
comprehensively, not just what crosses the network border. ... For a large
network, that can make troubleshooting and oversight more complex. “Something
like this is fine for small networks,” David explains, “but it certainly
complicates troubleshooting and oversight on larger networks.” David also sees
potential for SENSOR to expand beyond historical analysis by adding real-time
alerting. “The paper doesn’t describe whether the flow collectors can trigger
alarms for anomalies like rapidly spiking UDP traffic, which could indicate a
DDoS attack in progress. Adding real-time triggers like this would be a
valuable enhancement that makes SENSOR more operationally useful for network
teams.” ... “Finally, the approach is fragile. It relies on precise bridge and
firewall configurations to push traffic through the RouterOS stack, which
makes it sensitive to updates, misconfigurations, or hardware
changes. Network Segmentation Strategies for Hybrid Environments
It's not a simple feat to implement network segmentation. Network managers
must address network architectural issues, obtain tools and methodologies,
review and enact security policies, practices and protocols, and -- in many
cases -- overcome political obstacles. ... The goal of network segmentation is
to place the most mission-critical and sensitive resources and systems under
comprehensive security for a finite ecosystem of users. From a business
standpoint, it's equally critical to understand the business value of each
network asset and to gain support from users and management before segmenting.
... Divide the network segments logically into security segments based on
workload, whether on premises, cloud-based or within an extranet. For example,
if the Engineering department requires secure access to its product
configuration system, only that team would have access to the network segment
that contains the Engineering product configuration system. ... A third prong
of segmented network security enforcement in hybrid environments is user
identity management. Identity and access management (IAM) technology
identifies and tracks users at a granular level based on their authorization
credentials in on-premises networks but not on the cloud.
Convergence of AI and cybersecurity has truly transformed the CISO’s role
The most significant impact of AI in security at present is in automation and predictive analysis. Automation especially when enhanced with AI, such as integrating models like Copilot Security with tools like Microsoft Sentinel allows organisations to monitor thousands of indicators of compromise in milliseconds and receive instant assessments. ... The convergence of AI and cybersecurity has truly transformed the CISO’s role, especially post-pandemic when user locations and systems have become unpredictable. Traditionally, CISOs operated primarily as reactive defenders responding to alerts and attacks as they arose. Now, with AI-driven predictive analysis, we’re moving into a much more proactive space. CISOs are becoming strategic risk managers, able to anticipate threats and respond with advanced tools. ... Achieving real-time threat detection in the cloud through AI requires the integration of several foundational pillars that work in concert to address the complexity and speed of modern digital environments. At the heart of this approach is the adoption of a Zero Trust Architecture: rather than assuming implicit trust based on network perimeters, this model treats every access request whether to data, applications, or infrastructure as potentially hostile, enforcing strict verification and comprehensive compliance controls.Initial Access Brokers Selling Bundles, Privileges and More
"By the time a threat actor logs in using the access and privileged credentials
bought from a broker, a lot of the heavy lifting has already been done for them.
Therefore, it's not about if you're exposed, but whether you can respond before
the intrusion escalates." More than one attacker may use any given initial
access, either because the broker sells it to multiple customers, or because a
customer uses the access for one purpose - say, to steal data - then sells it on
to someone else, who perhaps monetizes their purchase by further ransacking data
and unleashing ransomware. "Organizations that unwittingly have their network
access posted for sale on initial access broker forums have already been
victimized once, and they are on their way to being victimized once again when
the buyer attacks," the report says. ... "Access brokers often create new local
or domain accounts, sometimes with elevated privileges, to maintain persistence
or allow easier access for buyers," says a recent report from cybersecurity firm
Kela. For detecting such activity, "unexpected new user accounts are a major red
flag." So too is "unusual login activity" to legitimate accounts that traces to
never-before-seen IP addresses, or repeat attempts that only belatedly succeed,
Kela said. "Watch for legitimate accounts doing unusual actions or accessing
resources they normally don't - these can be signs of account takeover."
