Archive for October, 2026

« Older Entries
Newer Entries »

26,400 working cloud and AI credentials exposed on Gitee & Cisco SD-WAN exploitation tied to Volt Typhoon

Posted in Commentary with tags on October 29, 2026 by itnerd

Independent researcher Ben Zimmermann scanned 7.25 million public Gitee repositories and found 26,400 cloud and AI credentials that still worked. They included thousands of Alibaba Cloud, DeepSeek, Zhipu BigModel and OpenAI keys, and he could not confirm that any had been revoked. More info here: 26,400 Valid Cloud and AI Credentials Found Exposed in Public Gitee Repositories

Separately, exploitation of a critical Cisco Catalyst SD-WAN flaw (CVE-2026-20127) has been tied to Volt Typhoon, with Lumen saying the group was exploiting the flaw by late 2025 and kept at it after Cisco disclosed it in February. More info here: The CISA adds eight Cisco SD-WAN flaws to KEV and gives organizations four days to fix them | The IT Nerd

Justin Beals, CEO & Founder, Strike Graph Said This:

On the exposed Gitee credentials:
“Zimmermann tested every key against the service that issued it, so 26,400 is a count of credentials that actually worked, and that makes this research more useful than most leak reports. He sent lists to the vendors with the most exposed keys, but nobody contacted the developers, and he couldn’t confirm that a single key had been revoked. If you ever pushed a key to a public Gitee repository, assume someone else still has it.

Nearly half of these credentials come from 2025, which matches how fast teams are building on AI right now. A developer grabs a DeepSeek or OpenAI key to get a prototype running, commits it, and moves on. Deleting it later doesn’t help, because the commit history keeps it and only the provider can revoke it. One of the exposed keys was an organization-admin credential for an enterprise AI platform. The two scanners agreed on only 1,001 of the keys, and each one caught about half the total. A clean result from one tool only tells you what that tool knows how to look for.

Every organization building with AI should inventory its machine credentials and give each one an owner, an expiration date, the narrowest scope that works, and a spend limit. Run more than one scanner, and revoke exposed keys at the provider instead of scrubbing the repo. As more agents move into production on keys like these, more companies will find their leaks in a cloud bill before they find them in their logs.”

On the Cisco SD-WAN exploitation tied to Volt Typhoon:
“Volt Typhoon keeps going after the network gear that sits between everything else, and SD-WAN fits that pattern. When an actor controls the device routing your traffic, they gain a view of everything moving across your network, and they blend in as part of the infrastructure. We saw the same playbook with Salt Typhoon inside the telecom carriers, and we are still working to get those actors out.

Most security programs still treat routers and edge appliances as plumbing. They get a configuration review once a year and a patch whenever the vendor ships one, while nation-state groups find the flaws first. For defense contractors and critical infrastructure operators, this goes beyond a security gap. If you can’t verify the integrity of the devices carrying your controlled information, you can’t honestly affirm that your controls work.

Network devices need to be first-class assets, with continuous configuration validation, log integrity checks, and a plan for the window before a patch exists. Expect more of this, because the edge keeps paying off for these groups.”

This is pretty bad and need urgent attention now. As in right now.

Leave a comment »

Canadian small business sales up for four months straight, a glimmer of hope after a tough 18 months

Posted in Commentary with tags on October 29, 2026 by itnerd

Xero today released its quarterly Xero Small Business Insights (XSBI) report, a snapshot of the health of the Canadian small business (SMB) sector based on actual aggregated and anonymized data from 12,000 Canadian small businesses using Xero, including sales performance, late payments, and time-to-be-paid.

The latest XSBI data shows Canadian small businesses had one of their better quarters in recent times, with sales up 2.0% year-over-year in the September quarter and four consecutive months of growth. While results remain well below the historical average, they offer early signs of recovery after a difficult 18 months, even as US trade tensions and elevated fuel prices continue to weigh on conditions. Payment times were largely unchanged at 10.5 days late on average (10.8 days in the June quarter), and Canada continues to have the longest late payment times of the countries tracked by XSBI.

Four consecutive months of sales growth

Canadian small business sales rose 2.0% year-over-year (y/y) in the September quarter, following an upwardly revised 0.4% y/y rise in the June quarter. It is the first time sales have grown for four consecutive months since April to July 2025.

Monthly results showed a 3.0% y/y rise in July, a 2.5% y/y rise in August and a 0.5% y/y rise in September. While these results remain well below the historical average of 4.0% y/y sales growth, they follow many months of declining sales and point to some improvement in performance.

Payment times little changed

Canadian small businesses were paid on average 10.5 days late in the September quarter, similar to the June quarter (10.8 days). Time to be paid, the length of time from when an invoice was issued to when it was paid, was 28.4 days, compared to 28.2 days in the June quarter, largely due to a 29.8-day result in September. Canada continues to have the longest late payment times of the countries tracked in XSBI.

Regional picture: Alberta leads as B.C. lags

Sales results for the September quarter were:

  • Alberta (+5.9% y/y) was the best performing of the regions tracked, benefitting most from higher global oil prices.
  • Ontario (+2.2% y/y) improved from the June quarter (+0.8% y/y), and could have been higher if not for a 1.7% y/y decline in September.
  • The Maritime Provinces (-1.4% y/y) had lower sales than a year ago, largely due to a 9.5% y/y fall in July.
  • British Columbia (-2.6% y/y) recorded lower sales than a year ago in every month of the quarter. However, B.C. remains the fastest place for small businesses to be paid (24.6 days).

The other provincial results on time to be paid for the September quarter were 28.0 days (Alberta), 28.9 days (Ontario) and 33.7 days (Maritime Provinces). Alberta’s result was pushed up by a rise to 34.1 days in September, its longest payment time since August 2023, which may be revised as more data becomes available.

Trade tensions hit exporter confidence hardest

The impact of renewed US trade tensions is falling hardest on small businesses that trade internationally. The Canadian Federation of Independent Business (CFIB) September Business Barometer found no statistical difference in 12-month confidence between exporters, importers and non-traders in September 2024. By September 2026, exporters reported a confidence score of 38.6 points, compared to 43.5 points for importers and 52.2 points for businesses not active internationally.

To access the latest data, visit xero.com/xerosbi.

Leave a comment »

Check Point’s new answer to AI-scale traffic

Posted in Commentary with tags on October 29, 2026 by itnerd

Check Point announced the Maestro Hyperscale Orchestrator (MHO) 200 today. It’s aimed at neoclouds, sovereign AI builds, telcos and large enterprises that can’t afford to choose between throughput and threat prevention.

  • Capacity quadruples. Fabric goes from 3.2 to 12.8 Tbps in the same 1RU box, with new 400GbE ports. Threat prevention scales to 4.5 Tbps per system, up from 1.1.
  • It addresses a common compromise. Check Point says many teams switch off advanced inspection or overbuy hardware to survive traffic spikes. This is built to avoid both.
  • It runs on top of the AI security push. It follows the AI Network Firewall, the NVIDIA collaboration on AI Data Center Security, and the agentic orchestration work released earlier this year.

IDC’s Pete Finalle puts the backdrop plainly: “AI is forcing a rethink of network capacity planning.”

It’s available now in all regions. The release is here.

Leave a comment »

FusionAuth 1.70 Lets Companies Securely Open Their Apps to Customers’ AI Agents

Posted in Commentary with tags on October 29, 2026 by itnerd

FusionAuth today announced FusionAuth 1.70, adding Client ID Metadata Document (CIMD) support and OAuth token exchange to help companies securely open their applications to customers’ AI agents.

AI agents give companies new ways to reach and serve customers, from sending invoices and buying tickets to ordering food and updating accounts. Yet 88% of organizations surveyed for FusionAuth’s 2026 State of AI and Identity report said AI deployment is outpacing their identity and security infrastructure. For companies building Model Context Protocol (MCP) servers, that raises two questions: which AI apps can connect, and what access they should receive.

FusionAuth 1.70 includes:

  • CIMD support: MCP-compatible AI apps can register automatically under rules the company sets so customers can sign in and approve access immediately instead of waiting on manual registration.
  • OAuth token exchange (RFC 8693): Services can request access to another system on customer’s behalf. Companies can configure policies to restrict the permissions granted, and FusionAuth records each exchange attempt and its policy decision.
  • Recorded intent, beyond RFC 8693: Services can include an optional intent statement with a token exchange request, and FusionAuth adds it to the exchange record for audit. Application providers can start agents with least-privilege access and review those records during testing to add only the permissions each agent needs.

CIMD support and OAuth token exchange are available today in FusionAuth 1.70 on the Enterprise plan. They run on the same FusionAuth platform that handles a company’s customer sign-ins, whether in FusionAuth Cloud, in the company’s own cloud, or in its own data center. FusionAuth doesn’t charge per token exchange. Documentation can be found at fusionauth.io/docs.

Leave a comment »

Guest Post: Put your zombie accounts to rest forever

Posted in Commentary with tags on October 29, 2026 by itnerd

More than 3 million email addresses found in data stolen by infostealers over the past year are no longer used by their owners. Some of them are abandoned but still active and pose a security risk. Millions more zombie accounts lurk on once-popular services like Groupon, dating apps people have ghosted, fitness services users gave up on, social media platforms, online shops, and even outdated work tools.

Key findings:

  • One year’s harvest. NordPass’ study examined 66.5 million unique email addresses harvested by infostealer malware over the past year and checked whether each address’s domain could still receive emails by checking publicly available DNS records.
  • Dangerous impersonators. The most concerning finding involves misspelled versions of major email providers. About 1% of all stolen addresses (664,000) are registered on live look-alike domains like “gamil.com.” Researchers add that some of those credentials may have been harvested when the person simply wanted to access their existing account — for example, on gmail.com — but accidentally typed “gnail.com” and logged in there. These domains work and receive emails. They’re just owned by someone else. In total, the researchers found more than 2,500 such look-alike domains mimicking Gmail, Hotmail, Outlook, Yahoo, and iCloud.
  • The dead can rise. Another 4% — 2.6 million addresses — are dead. The domains have expired, reject incoming emails, or never existed. But expired domains can be bought and re-registered by anyone, giving the new owner control over every account still linked to those addresses.
  • The harmless dead. The research also found that nearly one in 14 leaked credentials (7%, or 4.7 million addresses) come from throwaway inboxes, while another 2.1% use disposable email services like YOPmail or Mailinator. Both were built to be abandoned.

The issue extends well beyond email. A March 2025 Secure Data Recovery survey found that up to 40% of users hadn’t logged in to accounts on services like Pandora and Groupon for over a year. For many other services named in the survey, including Dropbox, Tumblr, Duolingo, Kickstarter, and Tinder, more than 30% of respondents admitted they had an account they hadn’t signed in to for at least 12 months.

Made for harvesting data

“Accounts created on look-alike domains are probably the most dangerous because those websites are essentially built to steal credentials — cybercriminals actively buy up common misspellings of famous brand names and expired domains to inherit their traffic and reputation, then use them for credential harvesting and intercepting emails intended for the original owner. After entering their details on such a fake page, users are typically shown an error or redirected to the real site — none the wiser. But their account data stays with the attacker,” says Karolis Arbaciauskas, head of product at NordPass and its parent organization Nord Security.

Other deserted accounts — on dating apps, photo services, and online shops, including AliExpress, Shein, and Temu — are also a security risk. The data shows that in the last couple of years, Western shoppers, including Americans, have been leaving Chinese retail marketplaces, but the number of accounts that were just abandoned and not properly deleted remains unknown.

“These zombie accounts stay online with your email address and password attached, and threat actors treat them as an easy source of data. Many are missing security updates and have no two-factor authentication, so attackers can break in with simple password-spraying attacks, testing a handful of common passwords across thousands of accounts. And since most of those abandoned accounts are linked to the owner’s main email address, the stakes are high,” says Arbačiauskas.

Most people reuse passwords or choose easy-to-guess combinations, so a compromised main email address can cascade into total account takeover. “If attackers get into your main email account, they can lock you out of everything else, steal your identity, drain your bank account, or hold your data for ransom,” says Arbačiauskas.

How to stay safe:

  • Delete unused accounts. To reduce the risk, review the services you’ve signed up for and delete your account if you no longer use the app, service, or platform.
  • Update the recovery email address. Check whether your old email addresses still exist and update the recovery email address on accounts that matter.
  • Never reuse passwords. Create strong, unique passwords for every account — a password manager can help you keep track of them. Otherwise, if one account gets broken into, all others are also at risk.
  • Use MFA. Whenever possible, enable multi-factor authentication for additional protection.

Leave a comment »

42 million documents exposed in GovTech platform leak

Posted in Commentary with tags on October 29, 2026 by itnerd


Cybersecurity researcher Jeremiah Fowler, a frequent collaborator of ExpressVPN, discovered an unprotected, unencrypted database linked to CitySourced, a civic engagement platform owned by Granicus. The platform allows residents to report issues to local governments in approximately 184 locations across North and Central America.

Jeremiah reviewed a limited sample of the exposed database, which included:

  • About 42 million documents, including 450,000 JSON files containing plaintext personally identifiable information (PII), such as names, addresses, email addresses, and phone numbers.
  • Images of resident complaints and code violation reports.
  • Thousands of internal records dating from 2016 to 2026, along with application directories and executable files.

You can read Jeremiah’s full findings here: https://www.expressvpn.com/research/citysourced-data-exposed

Leave a comment »

AppViewX Expands Quantum Trust Hub with Cryptographic Discovery, Posture Management, and Remediation

Posted in Commentary with tags on October 29, 2026 by itnerd

AppViewX today announced new capabilities for Quantum Trust Hub, now available as a standalone solution to help enterprises discover cryptographic assets, assess quantum-related vulnerabilities, prioritize risks by business impact, and take decisive action through integrated remediation workflows. The new capabilities include an agentless-first approach that leverages organizations’ existing security infrastructure to discover cryptographic assets without the additional overhead of agents and sensors. Organizations can also upload Cryptographic Bills of Materials (CBOMs) and Software Bills of Materials (SBOMs) from third-party tools to build a single, unified inventory of cryptographic assets, and initiate remediation planning and execution through existing ITSM solutions.

As enterprises prepare for the transition to post-quantum cryptography (PQC), understanding where vulnerable cryptography exists across their environments is becoming increasingly critical. NIST finalized its first PQC standards in 2024 and has outlined plans to begin phasing out RSA and ECC in 2030. By that same year, Executive Order 14412 directs federal agencies to transition high-value assets and high-impact systems to PQC for key establishment, with PQC adoption for digital signatures to follow by the end of 2031. Meeting these deadlines requires a comprehensive understanding of cryptographic exposure, yet traditional certificate inventories alone do not tell the full story. Cryptographic exposure extends beyond certificates to TLS configurations, keys, and application code, making it difficult to assess true risk and know where to focus first.

Addressing these challenges requires continuous visibility across multiple layers of cryptographic infrastructure. Quantum Trust Hub delivers this through a vendor-neutral approach, consolidating findings from existing security tools and third-party inventories. By building a unified cryptographic inventory and extending visibility across certificates, TLS configurations, cipher suites, and code libraries, the platform gives security teams the cryptographic posture management foundation they need, enabling continuous assessment of exposure, risk-based remediation prioritization, and the cryptographic agility to respond quickly as threats and standards evolve.

With the latest Quantum Trust Hub capabilities, enterprises can:

  • Expand cryptographic discovery using existing security investments: Through a new integration with Qualys, alongside existing Tenable support, organizations can discover cryptographic assets and configurations using security tools already deployed across their environments. AppViewX’s agentless-first approach reduces cost and complexity while supporting deeper discovery when needed.
  • Build a single, unified cryptographic inventory: CBOMs and SBOMs from third-party tools can be uploaded directly into the platform, consolidating findings from multiple discovery sources across certificates, TLS configurations, cipher suites, keys, and application code. This provides the foundation for true cryptographic posture management.
  • Prioritize what matters most: Assets are classified as quantum-vulnerable or quantum-resistant. Business application context and NIST-aligned severity ratings enable organizations to surface the most critical remediation priorities to address first. Executive dashboards and continuous monitoring track posture and migration progress over time.
  • Drive remediation through existing ITSM workflows: Cryptographic findings are sent directly to ServiceNow as ready-to-work tickets containing affected assets, identified risks, recommended fixes, and ownership information. Tickets are grouped by remediation scope, with status tracked end-to-end so teams can ensure findings translate into action, not just reports.
  • Move faster with cryptographic agility: By maintaining a comprehensive, continuously updated inventory of cryptographic assets and their risk status, organizations can respond quickly to new vulnerabilities, policy changes, and evolving PQC standards, without having to scramble to understand their exposure from scratch.

The expansion of Quantum Trust Hub reinforces AppViewX’s broader mission to secure machine and AI identities, credentials, and cryptography through a cohesive product portfolio as enterprises navigate the transition to post-quantum computing. Quantum Trust Hub works alongside AppViewX’s CLM and SSH Key Lifecycle Management solutions to accelerate remediation of certificates or SSH keys that require replacement as part of a PQC assessment. Additionally, AppViewX recently advanced its Agent Identity Security solution to discover, govern, and control the actions of AI agents at runtime while extending its core CLM and PKI products to issue cryptographically verifiable, quantum-resilient identities for AI agents, establishing an additional layer of trust as they access enterprise systems. Together, these offerings provide enterprises with a more comprehensive approach to securing machine and agent identities while managing cryptographic risk across their environments.

To learn more about Quantum Trust Hub, watch AppViewX’s on-demand webinar, “The Quantum Readiness Clock Is Ticking: What to Do Now” or explore a 30-day quantum readiness assessment by talking to an AppViewX expert.

Leave a comment »

PeopleSoft Flaw: Three Vulnerabilities, One Attack Chain

Posted in Commentary with tags on October 29, 2026 by itnerd

New research from Zero Day Initiative found that the actively exploited PeopleSoft vulnerability, CVE-2026-35273, actually encompasses three distinct vulnerabilities that attackers can chain together.

More info here: https://www.zerodayinitiative.com/blog/2026/10/20/cve-2026-35273-ssrf-to-rce-in-oracle-peoplesoft

Chris Jacob, Field CISO at Securonix said this:

“The danger of vulnerabilities like CVE-2026-35273 is how easily they allow threat actors to abuse internal trust relationships, transforming a critical enterprise application into a springboard for a much broader attack.

Once an attacker compromises a server, the battleground shifts entirely. The priority must immediately pivot from blocking the initial exploit to tracking post-compromise behavior, specifically how attackers attempt to establish persistence, harvest sensitive data, move laterally, or misuse trusted system credentials.

Security teams cannot rely on patching or perimeter controls alone, especially when attackers can mask their presence behind authenticated, seemingly legitimate local sessions. True resilience requires continuous, unified visibility across identities, applications, and infrastructure. Organizations need behavioral analytics that evaluate the full context of an action over time, allowing operations teams to distinguish routine administrative tasks from anomalous activity. The critical question isn’t just whether a vulnerability was triggered, but whether the trailing activity signals a deeper compromise and how fast defenders can isolate it.

I would spend some time to read this and take the action that is needed in order to save yourself some heartache.

Leave a comment »

Zoho Rebuilds Campaigns Platform With Simpler Email and SMS, Native WhatsApp and AI Agent Workflows

Posted in Commentary with tags on October 29, 2026 by itnerd

Zoho Corporation today announced a comprehensive update of Zoho Campaigns, its multichannel marketing platform. The update introduces a redesigned campaign experience, a template editor rebuilt from scratch, faster contact imports and email sends, native WhatsApp marketing, and support for AI agent workflows that can coordinate multi-step tasks through compatible AI tools.
                                                             
Together, these new features make everyday campaign work easier, from building an email and selecting an audience to coordinating messages across email, SMS and WhatsApp. Marketers can now create and manage campaigns with fewer steps, work with customer data across multiple channels, and use compatible AI agents to handle tasks ranging from campaign preparation to performance analysis.

The rebuilt Zoho Campaigns platform improves the speed of key campaign operations. In Zoho internal testing, sending 100,000 emails is now up to four times faster than before, and importing 100,000 contacts is now up to five times faster.

Despite these extensive updates, pricing for Zoho Campaigns remains the same, with paid plans starting at $4 a month.

Additions to Zoho Campaigns include:

A simpler way to run multichannel campaigns

  • WhatsApp (New): Bulk campaigns using approved message templates, trigger-based messages, and WhatsApp steps inside automated workflows.
  • Email: Fewer setup steps, clear recipient counts, and streamlined A/B testing across subject line, sender and content.
  • SMS: Simplified gateway setup, including a streamlined 10DLC brand and campaign registration process for US senders.
  • Frequency capping: Limits how often a contact is messaged across all three channels, so a customer in an email campaign and an SMS campaign is not contacted twice in the same window.

Connect compatible AI agents to campaign workflows

Zoho Campaigns now supports the Model Context Protocol (MCP), an open standard that allows compatible AI agents to connect to supported application capabilities.

Through supported integrations and configured permissions, marketers can use natural-language instructions to coordinate multi-step tasks in Zoho Campaigns. Rather than treating each action as a separate request, compatible agents can combine supported operations into a broader workflow.

For example, a marketer can instruct a compatible agent to:

  • Prepare a seasonal campaign: Build a Black Friday campaign for an existing customer segment, use existing HTML content, send a test and schedule delivery.
  • Reuse successful campaigns: Analyze the last 10 campaigns, identify a strong performer, clone it and adapt the content for a new promotion.
  • Prepare a re-engagement campaign: Identify contacts who have not engaged in the last 90 days, create a re-engagement segment, tag the contacts and prepare a campaign.

Depending on the connected agent, available capabilities and permissions, workflows can support campaign preparation, audience management, segmentation, performance analysis and other campaign operations.

The goal is to extend AI beyond content generation into campaign execution and coordination. Marketers remain responsible for campaign decisions, with review and approval requirements determined by the configured workflow and permissions.

AI-assisted content and more precise segmentation

Zoho Campaigns also improves the tools marketers use to create content and define audiences, helping marketers combine more relevant content with more precise audience targeting, while reducing repetitive work:

  • Zia Content Analyzer: Built into the template editor, this AI assistant checks spelling, grammar and style in real time, suggests tone, clarity and phrasing improvements, and drafts subject lines, preheaders and body copy for consistent email, SMS and WhatsApp campaigns.
  • Campaign Context: Apply a shared brief—including brand voice, offer details and audience information—across a defined set of campaigns to help maintain message consistency.
  • Advanced segmentation: Build audiences using layered criteria, engagement metrics and complex condition groups, with faster processing for large databases.

A template editor redesigned for everyday marketing

The new template editor is designed to help marketers create on-brand emails, adapt content for different devices and easy reuse designs across campaigns.

Key improvements:

  • Themes: Define primary and secondary brand colors once and apply them consistently to linked template elements.
  • Dedicated mobile editing: Work in a separate mobile-design view to adjust layouts for smaller screens.
  • Dynamic content: Personalize individual content blocks for different audience segments within the same email.
  • Reusable content blocks and centralized styles: Save frequently used components and manage shared brand settings in one place.
  • Cleaner HTML output: Generate more streamlined email markup, supporting consistent rendering across email clients.

Together, these changes reduce repetitive editing and help teams maintain consistent branding as they create and reuse campaigns.

A better campaign workspace

The rebuild also modernizes how marketers find and act on campaign data, reducing operational friction and helping marketers to move from campaign planning to execution more efficiently.

  • Redesigned workspace: A modernized interface with custom list views, advanced filtering and contextual information helps marketers find the data and actions they need.
  • Team collaboration: Shareable custom views and native contact tagging help teams organize and work with campaign data.

Unified reporting across channels

The updated reporting experience provides channel-level insights and a broader view of contact engagement, helping marketing teams connect campaign activity with contact engagement and business performance:

  • Channel dashboards: Access dedicated dashboards for email, SMS and WhatsApp, alongside a unified contact dashboard.
  • E-commerce integrations: Track ROI, revenue, orders and cart performance.
  • Zoho Analytics integration: Sync contact fields and campaign activity across email, SMS, WhatsApp and workflows for reporting in Zoho Analytics.
  • Unified contact scoring: View engagement across supported channels to understand contact activity in one place.
  • Centralized UTM management: Manage campaign tracking parameters more consistently.

Closer integration with CRM and other systems

Zoho Campaigns strengthens connections between marketing execution, customer data and external applications.

  • Live Zoho CRM sync: Keep contact data and audience segments up to date.
  • CRM custom views: Import contacts from a specific Zoho CRM custom view directly into a mailing list without recreating the filters.
  • Campaign creation from CRM data: Create campaigns using Zoho CRM data without switching between products for every step.
  • Expanded webhooks: Send campaign data to external applications to support connected workflows.

These integrations help reduce manual data handling and keep marketing activity better aligned with customer information in connected systems.

Pricing and availability

The updated Zoho Campaigns is available now in all regions, with pricing unchanged. SMS and WhatsApp marketing capabilities are included at no additional platform charge; sending messages requires purchased credits.

Zoho Campaigns continues to offer a free plan for up to 2,000 contacts and 6,000 emails per month.

For more information, visit www.zoho.com/campaigns

Leave a comment »

Cybernews spoke to 100 AI researchers. This is the biggest threat from AI

Posted in Commentary with tags on October 29, 2026 by itnerd

Cybernews has posted a new survey where they polled over 100 AI researchers on a wide range of AI topics, including their views on the pace of AI development, AI’s long-term impact, AGI and ASI timelines, and the greatest source of AI harm.

Their survey comes amid heightened concerns over AI’s existential risk, reignited by a public resignation of Anthropic researcher Jacob Coxon, who said that he and other employees in AI labs earnestly believe the technology can wipe out humanity by the end of the decade.

Their survey also indicates a widespread anxiety among the AI research community, as 30.4% of respondents say it is very plausible that a self-improving AI could pose a major risk to humanity within the next 5 years. Another 33.3% say that it’s somewhat plausible.

Other key findings:

  • Nearly 40% call for frontier AI to be slowed down, while 15.7% want it to be stopped completely.
  • More than four in ten (43.1%) believe there’s at least a 10% chance that AI will cause a catastrophic event this century.
  • 52.4% say a fundamental breakthrough is probably or definitely needed for AGI; 69.6% say the same for ASI.
  • When asked about the biggest source of AI harm over the next 10 years, 37.9% of researchers pointed to human misuse, such as cyberattacks, bioweapons, and disinformation.


Find many more insights by exploring the full report here: https://cybernews.com/ai-news/ai-researchers-survey/

Leave a comment »

« Older Entries
Newer Entries »