Compliance proves the past. Monitoring proves the present. ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­    ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏  ͏ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­ ­  
View in browser
Logo

06-ai-at-scale-globe

From our CEO…

The conversation around responsible AI has gathered pace following the hacking incidents reported by several major players over the summer.

 

One of the most significant developments came in early October with the announcement of the US Super Intelligence Force, a new federal task force to coordinate the US government's approach to artificial intelligence. Its aim is to balance America's ambition to maintain global leadership in AI with concerns about the technology's risks. It's a sign that responsible deployment is being taken more seriously, although it remains to be seen what oversight, if any, it will recommend.

 

You can read my reaction in TechRadarPro which outlines my view that the most effective form of AI oversight is a broad, diverse set of AI companies holding each other accountable. This is more achievable than a single international body and would foster a stronger sense of collective responsibility. Rather than answering to an outside organization, companies within the sector would have greater impetus to shape the direction of AI and to manage risks as they emerge. Regulation has its place in setting parameters and minimum standards, but it cannot realistically keep pace with the speed of change.

 

That's why continuous monitoring must sit at the heart of AI governance. 'Point-in-time' checks are not enough for a technology that can circumvent safeguards, act autonomously, and is becoming more independent. There is now broad agreement that AI deployment needs to be managed well. But with a technology gaining capability by the day, we cannot afford to spend too long deliberating before we act.

 

–   Nikolas Kairinos, CEO of RAIDS AI

Introducing the RAIDS Browser Extension
Know when the AI you use goes rogue

How many AI answers did you rely on today? And how many did you actually check?

AI now drafts our emails and summarizes our reports, and the more useful it gets, the less we stop to question it.

But AI rarely fails loudly; it drifts. You ask for a report's key findings and the summary dwells on the background. You ask it to compare options and it quietly skips one. These answers sound confident and read well, so they're easy to miss, until you've already sent or acted on them.

That's why RAIDS is coming to your browser.

We're starting with ChatGPT, checking every answer as it arrives. A blue dot means everything looks normal. A red dot means something's off, and one click shows you the full breakdown. Everything happens right in your browser, from signing up to viewing alerts, tracking your usage, and giving feedback that helps it improve over time.

Less second-guessing. More insight and control over the AI you rely on.

The free beta launches first on the Microsoft Edge Add-ons Store, with 10,000 credits to use until they run out or the beta ends. More AI tools are on the way, along with a version for companies.

Get notified at launch

Monitoring the risks of AI in healthcare

Screen Shot 2026-11-03 at 02.50.56.135 PM

AI use is growing in almost every sector, but in some, the risks of rogue AI are far greater than in others. Healthcare falls firmly into that category.

Our CEO, Nik Kairinos, recently wrote an article for Healthcare Today exploring how AI is moving beyond administrative tasks into clinical decision-making and treatment delivery, and the new challenges this creates for ensuring its reliability. While AI can improve diagnosis, increase capacity, and enhance patient care, its performance can change after deployment due to evolving data, patient populations, and clinical practices. Pre-deployment testing alone is therefore not enough to manage the risks.

Regulators increasingly recognize the need for ongoing oversight, with the EU AI Act, UK medical device regulations, and ISO 42001 all reinforcing the importance of monitoring and risk management.

Continuous, independent monitoring can help healthcare organizations detect changes in AI behavior, identify emerging bias, flag unexpected outputs, and maintain audit trails. While it cannot replace clinical oversight, it adds an extra layer of assurance.

Read the article

The EU AI Act – the next milestone approaches

raids-ai-regulation-v2

We are fast approaching the next EU AI Act milestone on December 2, 2026. From that date, the Act prohibits AI systems used to generate harmful content, including non-consensual intimate imagery and child sexual abuse material. The date also marks the end of the transition period for existing generative AI systems. These systems must now make AI-generated text, images, audio, and video identifiable through machine-readable marking.

These are critical changes that will make a tangible difference to the responsible deployment of AI, and it’s important that organizations understand their obligations. However, the milestone is also a reminder that the EU pushed back deadlines as part of the Digital Omnibus, and that progress toward AI regulation remains painfully slow.

 

Our new whitepaper with Asia-ISAC 

Screen Shot 2026-11-03 at 03.22.46.826 PM-1

We’re pleased to share a new whitepaper created in collaboration with Asia-ISAC (Asia Information Sharing and Analysis Center), the first and only cross-industry, non-profit regional hub dedicated to helping companies unlock growth through safe and secure AI adoption while reducing the cost and impact of cyberattacks across Asia. The paper explores how organizations can manage AI risks as they move from pilot projects into live production, and how to close the Governance Latency Gap: the critical delay between a model’s initial certification and its often unpredictable behavior once it goes live.

It proposes a lifecycle approach that combines governance design, real-time monitoring, incident response, and ongoing compliance assurance, with continuous, independent monitoring to identify emerging risks and create a persistent audit trail. Drawing on examples from financial services, logistics, and enterprise AI adoption, it also examines Asia’s fragmented regulatory landscape and the importance of human oversight. Ultimately, it shows that effective AI governance requires technical monitoring, institutional accountability, and human judgment working together.

Read the whitepaper

MEDIA COVERAGE

RAIDS AI in the press

Our CEO, Nik Kairinos, has been offering his views across a range of publications and sectors. Here are some of the latest highlights:

UKTN-logo-web

✅ Nik explains why AI monitoring should be part of investors' due diligence, in an article for UKTN.

Techerati-Site-Images-1000-x-500px-2026-09-25T150339.131

🇦🇺 In September, reports emerged that an OpenAI agent had accessed Australian government files in June, but the incident wasn’t reported to authorities until nearly three months later. In Techerati’s coverage, Nik highlighted the importance of continuous monitoring to detect and document unexpected AI behavior as it happens.

techtarget

💻 TechTarget considered what the AI safety fallout might mean for CISOs, and Nik offered practical advice for responsible implementation.

TechnologyMagazine

🚨 Reflecting on escalating AI risks, Technology Magazine included Nik’s thoughts on what industry safeguards could look like. 

Connect with us…

We’re always keen to connect with others working in and around AI.

 

You can follow RAIDS, or connect with Nik, Franki and the team on LinkedIn to stay up to date with what we’re building and thinking about.

 

Our RAIDS Intelligence Hub is a collection of content including documents, reports and videos, providing everything you need to understand AI governance, regulatory compliance, and how RAIDS provides continuous behavioral monitoring for enterprise AI systems. 

 

You can also subscribe to make sure you receive future newsletters here.

 

And if you’d like to talk about AI risk, safety, or how RAIDS could support your organization, contact our expert team today.

Logo
LinkedIn
website icon

RAIDS AI, 26 Nikis Avenue, Nicosia, 1086, Cyprus

Unsubscribe Manage preferences