Fraud is no longer only one stolen card or one fake account. It is synthetic identities, account farms, bots, mule networks, bonus abuse, phishing, payment manipulation, chargebacks, fake leads and account takeovers popping up across thousands of sessions at the same time.

By the time a human analyst sees the full picture, the fraudster might already be gone, off to the next thing.

��

Why Humans Are Hitting A Limit

��

A good fraud analyst is still valuable. Very valuable, actually. Humans get the bigger picture, the business nuance, customer behaviour, and the awkward edge cases better than any model by itself. The issue is scale.

An analyst can review a case. Modern scoring engines can evaluate large numbers of signals in real time, while AI-assisted analytics help improve detection quality over time. Device history, IP geolocation, behavioural patterns, transaction velocity, account links, session timing, email structure, document signals, wallet behaviour, previous fraud clusters all at once.

This is where modern becomes useful: not because it mysteriously knows who is guilty, but because it can rank risk across gigantic volumes of activity before a human team even realises where to look.

��

What AI Sees That People Miss

��

A single fresh device may be normal. A fresh device plus copied input, proxy use, repeated browser configuration, similar account creation time, and links to old bad users? That seems different, really.

AI is strong because it reads combinations. Human analysts often need a very clear trigger: suspicious payment, failed sign in, refund abuse or chargeback. The scoring engine can identify elevated risk before visible fraud occurs.

Instead of producing a black-box verdict, modern fraud platforms can show exactly which triggers contributed to the risk score. Analysts can review activated rules, device links, behavioural anomalies, and transaction context, making investigations faster and easier to validate.

Rather than relying on a single signal, risk decisions are often built from multiple factors working together. Visibility into those factors helps teams improve policies, reduce false positives and respond faster to emerging threats.

Why Speed Matters So Much

��

Fraud teams used to look mostly at what happens right after the transaction. But now it is not enough. In a lot of businesses, the dangerous moment shows up earlier.

In TransUnion’s 2024 fraud report they said 13.5% of global digital account creation transactions in 2023 were flagged as suspected digital fraud, so account creation became one of the riskiest steps in the customer journey. That figure explains why the playbook changed. If you wait until the first payment abuse, or the chargeback shows up then too late.

Modern AI systems can track the entire journey and continuously refresh risk ratings while the user is moving through it. A legitimate user glides along. A suspicious one gets slowed down, questioned, throttled or redirected to a human case review before the meaningful harm actually starts.

��

Why This Matters For The Business Side

��

Fraud detection is not only a security job now. It touches revenue, customer experience, regulatory compliance, and investor confidence

A slow fraud process can hit a company in a bunch of ways:

• Real fraud gets through before anyone reacts
• Good users get blocked by blunt rules
• Support teams drown in manual reviews
• Withdrawals, deposits or onboarding become slower
• Risk teams spend too much time on low-value alerts

Analysts should not spend their days clicking through obvious junk. Their time is better spent on complicated cases, improving controls, checking edge decisions, and hunting down new fraud methods that keep changing. Otherwise the system stays noisy while the real signals stay hidden.

��

AI Also Protects Good Users

��

Here is the bit people often miss: stronger fraud detection is not just about denying more accounts. It is also about denying fewer legitimate ones. AI can add context, it can tell when a new device is not meaningfully risky, when a quick transaction really matches normal history, or when a user looks unusual but still not dangerous.

That helps reduce false positives, and false positives matter a lot. They lead to support tickets, lost revenue, and angry customers.

put the global average breach cost at $4.4 million, which shows how costly weak security and slow reaction time can get. Faster detection and containment were a big reason the average dropped from the year before, because the damage just stacks up when incident response lags a little too much.

��

Human Analysts Are Not Going Away

��

The best fraud squads use AI to sift through the noise and human expertise to handle the thorny items. Analysts look into model alerts, inspect odd signals, tune the rules, double check conclusions and map what’s happening back to real business context.

This partnership really matters. Purely automated blocking, without any oversight, can turn risky. It might drive unfair outcomes, fail to catch freshly emerging fraud logic or weaken user confidence.

��

Why This Matters Right Now

��

Fraudsters already run automation, and they tend to adapt quicker than manual teams can respond, especially when volumes spike. So the main question becomes, how well companies run the system: clean data, models you can explain, human review in the loop, and feedback cycles that get sharper over time.

AI isn’t exactly replacing fraud analysts, it’s shifting the whole job, if you look close.

There is less of the manual searching, more deep investigation, less repetitive checking. For modern organisations this change is kind of huge, because fraud doesn’t wait in a neat queue anymore, it just keeps moving.

The post AI Is Now Better At Catching Fraudsters Than Human Analysts: Here’s Why That Matters appeared first on 91̽��.

When people think of AI they often think chatbots and image generators. European businesses are using the technology for something less visible. They are using it for ocean mapping, carbon management and environmental monitoring projects that once demanded huge amounts of manual labour.

Sasha Rubel, Head of Public Policy for Generative AI at AWS, believes the discussion has entered a different stage. She said, “The conversation around AI and climate is shifting in an important way. We need to move beyond the question of whether technology can help, and focus on how we use it to address the defining challenge of our time: preserving the planet.”

Her comments come as businesses look for tools that can better help turn environmental commitments into day to day activity.

��

What Is AI Already Doing In Europe?

��

Talk about AI often speaks of future possibilities, but Rubel’s examples are already happening now:

“Across Europe, AI is already being deployed to address some of our most difficult environmental challenges. Carbon is being mineralised into building materials by companies like Paebbl. Zero-emission vessels from startups such as XOCEAN are mapping our oceans at scale. These are examples of AI helping in the present, on operational infrastructure, delivering measurable results.”

The projects come from different areas of the economy. One turns carbon into construction materials. A second sends autonomous vessels into the sea to gather environmental data without producing emissions.

Neither project looks anything like the consumer AI products that dominate tech news. Their value comes from gathering information, analysing conditions and helping organisations make environmental decisions faster than traditional methods allowed.

Rubel sees that work as one tool available to Europe. She said, “But the opportunity will only be realised if businesses treat sustainability as a design principle, not a reporting obligation. This means embedding environmental accountability into every architectural decision, every procurement choice, every product roadmap. Europe has set its Net-Zero targets. If we stay committed, and ensure that AI becomes one of the defining tools of our climate response, we can close the gap between ambition and action.”

��

��

What About The Environmental Cost Of Technology?

��

World Environment Day also brings up a less discussed fact. The devices people use every day come with environmental costs long before they are in our hands.

Arjen Steenbergen, ESG Manager at Trust International, says many discussions concentrate on disposal, when manufacturing often creates the biggest footprint.

He said, “World Environment Day is a timely annual check-in reminding us that meaningful climate action is rarely driven by a single breakthrough. More often, it is the result of incremental improvements made consistently over time.”

More than 80% of a headset’s climate footprint can be generated during manufacturing. Global e-waste is expected to exceed 80 million tonnes by 2030.

Steenbergen said, “The challenge facing the electronics industry highlights exactly why this approach matters. More than 80% of a headset’s climate footprint can be generated during manufacturing, while global e-waste is expected to surpass 80 million tonnes by 2030. These figures show that sustainability cannot be treated as an end-of-life issue but as a concern at every stage of a product’s lifecycle, from sourcing materials and manufacturing to packaging, use, and eventual disposal.”

��

How Can Small Decisions Make A Difference?

��

Trust International says environmental work often comes down to hundreds of choices that consumers don’t even notice.

Steenbergen said, “At Trust, we have focused on every decision regarding the creation, sourcing and eventual clearance of our products. Over the past year, this has included increasing the use of recycled materials, reducing plastic and foam packaging by 25% and 32% respectively, and continuing to strengthen the standards and certifications that help validate our progress. Achievements such as maintaining EcoVadis Gold status for five consecutive years demonstrate the value of turning ambition into measurable action.”

Those changes may not sound as bad when viewed individually. Packaging materials, sourcing decisions and product certifications do not usually generate front page coverage. Environmental progress often comes from work of that kind.

Steenbergen believes businesses will ultimately be judged on evidence instead of just promises.

He said, “As scrutiny around environmental performance continues to grow, businesses will increasingly be judged not by the sustainability targets they announce, but by the progress they can demonstrate. Independent certifications and transparent reporting play an important role in moving the needle for the industry to deliver meaningful results and, most importantly, for the betterment of the earth.”

That thought feels especially relevant on this year’s World Environment Day. The conversations about climate are often around future targets but the examples from AI developers, ocean mapping companies, carbon technology businesses and electronics manufacturers look more at work that’s already in progress.

The post World Environment Day 2026: How Tech Is Impacting Europe’s Climate Footprint appeared first on 91̽��.

That figure has grown faster than most organisations anticipated; the governance infrastructure around those deployments hasn’t kept pace, and security budgets remain sized for a slower rollout. The reality is that Claude is already embedded across many enterprise environments with limited visibility into how it’s being used.

at no cost for organisations running Claude across their business – including Claude AI, Claude Code and Claude Cowork.

The paid platform continues to offer the full advanced capability set; the free tier is designed to give any organisation a functional governance starting point with minimal setup time.

��

What The Free Tier Covers

��

The no-cost offering is focused on discovery and visibility: user, model and agent discovery, agent access graphs, tool call visibility, token usage and cost tracking, and conversation threads. The intent is to give security and IT teams a clear baseline picture of Claude activity across the business before moving to policy enforcement or deeper controls.

“Deploying Claude is the starting point,” says Pakshi Rajan, Chief GenAI and Product Officer at Portal26. “What organisations need upfront is the infrastructure to discover all Claude AI, Claude Code, and Claude Cowork usage, surface all conversations and tool calls, govern it, protect it, and . Portal26 is the only platform that provides foundational capabilities free of cost, making security extremely accessible and high-value enterprise Claude deployments a reality.”

In its April 2026 blog post, Anthropic acknowledged that while it has begun to provide some oversight tooling, ultimate responsibility for governance and controls rests with the deploying organisation.

According to Portal26, the primary security exposure exists in the shortfall between Anthropic’s out-of-the-box capabilities and true enterprise security demands.

��

Advanced Capabilities And The Broader Platform

��

Beyond the free tier, the fully integrated platform adds: comprehensive security and risk detection, real-time security policy enforcement, token policy and cost enforcement, MCP controls and policy enforcement, enterprise integrations into IDP, SSO, SIEM and incident response platforms, and access and privacy controls. These are available at Portal26’s standard pricing.

The platform also addresses – undiscovered AI tool usage outside sanctioned channels. Portal26 claims organisations using the platform detect three times more Shadow AI than those relying on legacy security providers, with ten times more security coverage overall. For enterprises where agent usage is expanding faster than IT visibility, this detection layer has become the starting point for any serious governance programme.

“Portal26 gives organisations the full-lifecycle AI management capability they need to move from cautious experimentation to confident, scalable, and measurable Claude adoption – securely, and at speed,” says Arti Raman, CEO of Portal26. “Most importantly, Portal26 provides critical governance and security capabilities for Claude free of charge, making it widely available and accessible to Claude users everywhere.”

The free AI governance and security offering is available now at

The post Claude Is Surging Across Enterprise – Portal26 Just Made Governance Free appeared first on 91̽��.

Fast forward to 2026, and the conversation has completely evolved. We are seeing the rise of a new breed of growth partners that are changing the rules entirely by building their entire business model around automation from day one.

For instance, forward-thinking agencies like as well as various others show how deeply automated intelligence can transform online advertising from a slow process of guessing to a quick and flexible growth engine. This paradigm shift revolutionises the way British businesses communicate with their customers and maximise their investment in online ads.

Because these agencies treat machine learning as foundational infrastructure rather than an optional add-on software program, they look less like legacy marketing shops and more like agile technology partners. They focus on total systemic overhaul, stripping away the slow processes that used to hold brands back.

��

Replacing Legacy Billable Hours With Agile Operational Infrastructure

��

Traditional agencies rely heavily on billable hours and massive human teams working to manage distinct fragments of a campaign. When consumer habits change rapidly, that old layout feels slow and heavy. AI-first agencies reshape this dynamic by shifting from a traditional pyramid structure to an agile diamond shape.

In this new model, routine administrative tasks, manual account setups and surface-level data collection are fully handled by automated systems. Because their internal technical systems are deeply automated, a small, highly specialised team can manage massive output volumes that used to require a cast of dozens.

This structural change allows mid-market businesses to access enterprise-grade precision without the historic overhead costs. Clients spend less time paying for basic manual labour and far more time getting direct access to senior strategists who can make fast, impactful commercial decisions based on real-time campaign performance.

��

Tracking Clean First-Party Data Signals Over Creative Guesswork

��

With the ongoing deprecation of third-party cookies and rising global privacy compliance requirements, brands can no longer rely on external platform metrics to guess who their buyers are. AI-first agencies reshape the creative process by feeding it high-quality, first-party data signals. They spend a massive amount of time building robust server-side tracking, CRM integrations and advanced attribution models for their clients.

According to an industry market analysis published by The MTM Agency, UK digital ad spend is expected to reach nearly £50 billion by the end of 2026, heavily driven by programmatic sophistication, connected TV, and retail media.

However, their research also notes that 56% of industry leaders cite AI and automation as top operational challenges due to transparency and algorithmic tracking concerns.

AI-first partners tackle exactly this challenge by paying attention to data sanitisation. They ensure that all automated bidding works in favour of the real business value and not some vanity metrics such as clicks.

Transforming Marketers From Production Workers Into System Architects

��

Because autonomous systems can handle repetitive tasks like basic keyword research, routine A/B testing, and manual campaign optimisation, the day-to-day role of a marketer is transforming. The industry is moving away from manual content production workflows and leaning heavily into what experts call agentic workflows.

The most effective marketing professionals are no longer those who will spend their whole day laboriously typing out their own copy or using stale old spreadsheets. The best agencies use people to be the system architects who design, control, and interrogate the AI models used to run the campaigns.

As detailed by insights from The MTM Agency’s Paid Media Summit, speakers emphasised that AI is now responsible for much of what historically required human optimisation, such as campaign setup, creative iteration, and cross-channel bidding. This shift places new importance on the inputs we provide platforms.

Rather than button-pressing, the role of paid media managers is shifting toward strategic direction, creative excellence, and model training.

Human oversight remains entirely vital to provide real-world commercial judgment, protect brand voice, and ensure strict data governance, but the sheer leverage that AI provides means that creative professionals can spend less time on administrative execution and far more time on high-level business strategy.

��

Scaling Personalisation Through Real-Time Behavioural Data

��

For decades, marketers grouped people into broad demographics like women aged 25 to 34 in London. In 2026, those broad buckets are no longer enough. UK consumers expect brand communications to feel like genuine, one-to-one conversations. AI-first agencies solve this by using autonomous systems to evaluate hundreds of distinct data signals in real time.

They look at immediate browsing behaviours, historical purchase habits, and even external elements like local weather patterns to serve a highly tailored message at the exact moment of intent. Managing millions of individual customer touchpoints across multiple digital platforms is a human impossibility, but it is precisely where machine learning thrives.

Automation frameworks make sure each and every outreach is completely personalised, ensuring that profit margins are protected while increasing conversions. Through the instantaneous linking of disconnected data points, these agencies enable brands to be there with the solution before consumers even know they’re ready to make a purchase decision.

��

Reshaping Discovery Frameworks For AI Search Engines

��

The way people discover information online is completely changing. We are no longer living in a world where search is strictly about typing a few disjointed keywords into a standard search box and clicking a list of blue links. UK consumers are now moving fluidly across conversational AI chatbots, voice prompts, and visual search tools.

Because search has become conversational, queries are longer and far more contextual.

People are looking for direct answers, clear comparisons and immediate reassurances.

AI-first agencies are leading this charge by analysing how large language models attribute data, ensuring their clients’ digital footprints are structured so that AI engines naturally pull, trust and recommend them within direct conversational answers.

They focus on semantic relevance, structured schema markup, and clear information accessibility so that a brand’s data is easily read and cited by conversational search models.

What is happening in the UK marketing environment is neither a transient phenomenon nor a mere change in technology. The whole essence of the way businesses evolve is being reinvented by agencies that have restructured their entire business processes through AI.

It has become increasingly apparent for brands in Britain hoping to thrive in the marketplace that success is no longer based on the size of the office or the number of junior employees who would be working for each individual account. Instead, it depends on the level of incorporation of artificial intelligence in their business activities.

The post How AI-First Agencies Are Reshaping UK Marketing In 2026 appeared first on 91̽��.

The disruption came one day after Anthropic filed paperwork for what is expected to be one of the biggest public listings in the technology sector. Founded in 2021 by former OpenAI employees, the company has become one of the most valuable names in AI.

��

Win Some, Lose Some: Why Is This Bad Timing?

��

The Register reported that a funding round in May valued Anthropic at around $965 billion (£717 billion). The publication also reported that the company could soon report its first quarter of operating profit.

Anthropic later said, “Earlier today, some users may have experienced intermittent issues or slower response times across Claude Code, Cowork, Claude.ai, and the API. Service has been fully restored, and we’re grateful to our users for their patience. Customers accessing Claude through Google Cloud’s Vertex AI or Amazon Bedrock were not affected.”

The disruption was little more than an inconvenience for many. For businesses using AI throughout daily operations, the outage was as an example of what can happen when a digital service becomes unavailable without warning.

��

Have Companies Built Too Much Around AI Tools?

��

Philip Miller, AI Strategist at Progress Software, believes many organisations have spent considerable energy adopting AI tools and far less energy preparing for service failures.

He said, “The Claude outage is a warning shot for enterprise AI. Not because Claude failed, but because many businesses have not designed for AI failure. The bigger question is what happens when your business has quietly started to depend on AI for research, coding, customer service, compliance, content creation, decision support, or workflow automation, and that AI suddenly stops responding. That is when AI moves from ‘productivity tool’ to operational risk.”

AI is everywhere – many organisations started using AI because it helped staff complete work more efficiently. Over months and years, those systems became woven into normal business activity and workplace routines.

When outages happen, businesses can discover that employees have built entire workflows around services controlled by external providers.

��

��

What Should Happen When AI Becomes Unavailable?

��

Miller said, “This is the part of enterprise AI that still does not get enough attention. Reliability is not just model uptime. It is the ability to keep the business operating when a model becomes unavailable. AI systems need governed context, fallback paths, policy controls, human escalation, explainable outputs, and a record of what happened.”

That view moves the conversation away from model performance and into business continuity. A company may have access to an advanced AI model, but work can slow down quite a bit when staff have no alternative systems available.

Things like human oversight and backup processes become valuable when automated services stop responding. Organisations that prepare for outages can continue working with less disruption.

Tech outages are nothing new and businesses have spent years preparing for interruptions. Miller believes AI deserves the same level of planning.

��

Is AI Now Being Treated Like Critical Infrastructure?

��

The timing of the outage generated discussion because it happened as Anthropic moved closer to a public listing that could rank among the biggest technology flotations in recent years.

According to The Register, Anthropic’s valuation now exceeds that of OpenAI. The publication also reported that Claude Code has strengthened Anthropic’s reputation with software developers.

For Miller, the outage demonstrated an important reality about how businesses now view AI. He said, “The timing of this outage, coming alongside reports of Anthropic moving toward a potentially enormous public listing, is a useful reminder.

“The AI market is being valued like critical infrastructure. Enterprises now need to implement it like critical infrastructure.”

The outage lasted only a few hours, but the event showed that any digital service can become unavailable and it’s up to businesses to stay prepared for any future ones.

The post Anthropic’s Claude Outage: Are Businesses Built Around AI Prepared For Failures? appeared first on 91̽��.

The data makes the case – women-only founded startups received 1.2% of VC funding in MENA in 2024, up from 0.47% in 2023. August 2025 broke the mould: two deals alone channelled $72.3 million to female-led ventures in a single month.

One was Gathern, a Saudi AI-powered vacation rental platform founded and led by Hala Aldosari, which raised $72 million in a round led by PIF’s Sanabil Investments, reaching a $266 million valuation. That makes Gathern the highest-valued in MENA founded and led by a woman.

The other deal was Phys, a Saudi youth health platform co-founded by Zainab Alshaber, focused on gamified activity tracking for children. Different sector, same indicator – the trend is spreading.

��

The Founders Building It

��

Nour Taher is targeting a major blind spot in the . As CEO of intella, a Saudi-based company originally founded in Egypt, she’s leading the development of Arabic-first AI – speech-to-text, natural language processing and analytics that work across 25 Arabic dialects. The company raised a $12.5 million Series A in 2025, led by Prosus, bringing total funding to $16.9 million.

The company is now launching Ziila, an Arabic digital human designed for enterprise customer engagement. With Arabic-first AI wide open globally, Taher is the closest to dominating the sector.

In the UAE, CozmoX is showcasing a completely different type of AI venture worth watching. Co-founded by Nuha Hashem, the company builds AI employees capable of doing the work of 100 people: customer service, sales, debt collection, audits, invoice processing. Only five months into its launch by early 2025, CozmoX had onboarded more than 50 companies in the Emirates.

This rapid growth proves both the product’s value and market demand – the UAE’s regulatory infrastructure and AI-first national strategy have created conditions that compress the distance between launch and traction.

Over in Egypt, Dr. Rasha Rady and Dr. Doaa Aref have turned Chefaa into one of the most interesting female-led healthtech startups to watch – an AI-powered e-pharmacy platform with GPS-enabled services now operating across eight Saudi cities, having raised $5.25 million from Newtown Partners and Global Brain.

��

��

What Changed To Make This Possible

��

Several changes aligned simultaneously to create this momentum. The UAE’s National AI Strategy 2031, Saudi Arabia’s PAISDA and the active involvement of both the UAE Ministry of AI and the Saudi Authority for Data and AI in backing women-led AI startups created a structural tailwind that has fundamentally changed what’s possible for founders in the region. Regulatory sandboxes, 100% foreign ownership rules and dedicated startup infrastructure have made the region considerably easier to build in than it was at the start of the decade.

Capital followed the strategy. AI startup funding in MENA rose 22% in 2025, with more than 60% directed to the UAE, and between 2022 and 2024 the region saw $660 million raised across 322 AI deals, with one in five VC deals in 2024 involving an AI company. Saudi Arabia led MENA VC in 2025 with $1.72 billion raised, up 145% year on year – that surge in capital helps serious startups secure funding, including female-founded ventures.

Dedicated programmes filled in the spaces that capital alone doesn’t cover. Google’s Women in AI Growth Academy, launched in 2024 as a partnership between Google for Startups, the Saudi Authority for Data and AI and the UAE Ministry of AI, offered an equity-free development programme to ten women-led in its first cohort.

The programme is designed precisely to address the pre-seed and seed stage where women-led companies have historically been most disadvantaged.

��

What The Next Generation Can Expect

��

The upward trend is clear and unmistakable. Government strategy, international capital and dedicated programmes have created a foundation that simply didn’t exist five years ago, and the founders who moved early are now demonstrating what’s possible at large. Gathern, intella and CozmoX are the proof of concept that the next generation can point to.

Gathern’s $266 million valuation, intella’s Prosus-backed Series A and CozmoX’s rapid UAE adoption all demonstrate something that wasn’t demonstrable three years ago: women-led AI companies in can reach serious scale with serious investors. Female entrepreneurs in the region are projected to double by 2030. The infrastructure – government strategy, VC investment, dedicated programmes – is more developed than it’s ever been.

The opportunity that remains most open is also the most distinctive: Arabic-first AI. Global models still underperform significantly on Arabic language tasks. The founders who build locally grounded, culturally specific AI tools for a 400 million-person Arabic-speaking market have a true first-mover advantage that no Silicon Valley lab can replicate from a distance.

Rather than a niche market, this represents one of the most underexplored segments in global AI development. Currently, women like Nour Taher are best positioned to capture it.

The post The Women Behind MENA’s Most Exciting AI Startups appeared first on 91̽��.

The Gulf region provides a compelling response to that. Saudi Arabia, the UAE and Qatar are in the middle of some of the most ambitious infrastructure buildouts in the world – dense new city districts, automated port and logistics hubs, advanced manufacturing zones, AI-powered hospital networks and utility infrastructure designed from the ground up for intelligent automation. These are new projects built with AI at their core, avoiding the need to retrofit outdated systems. According to GCC states treat AI as a core strategic capability rather than a trial, which changes the procurement dynamics.

Sarah Rees, CEO of Signwl, which tracks GPU cloud pricing and AI infrastructure capacity globally, points to an unusual indicator in the current data. “MENA is the tightest major GPU region we track in terms of capacity,” she says. “H200 spot pricing is currently trading 5% above on-demand price, which suggests material utilisation relative to supply – a highly unusual occurrence.

MENA tech founders also have local access to frontier silicon including Nvidia’s Blackwell GB200, which vision-language-action models and robotics depend on.” She also notes a concentration risk: 85% of GPU SKUs available across MENA cloud regions are Nvidia, compared to 62% in the US and 65% in Europe, which she says founders should build around now to maintain silicon flexibility.

��

Which Sectors Feel It First

��

Logistics and ports offer the most immediate potential for development.

The Gulf has been automating container terminals for years and the arrival of production-ready gives operators what they’ve been waiting for: the ability to close the distance between ambition and execution. Elmer Morales, a two-time founder and former big tech engineering leader currently building AI-native software infrastructure, is direct about this. “Jebel Ali is an obvious candidate. So is NEOM’s logistics backbone,” he says. “These are environments with controlled variables, high transaction volume, and clear ROI on automation. The infrastructure appetite is already there.”

Healthcare is the sector Morales describes as the “sleeper.” The Gulf has invested heavily in hospital infrastructure but faces a persistent shortage of skilled labour in diagnostic and operational roles. “Physical AI agents that assist in triage, inventory, and patient flow are not science fiction at this point,” he says. “They are procurement decisions.” Rees adds a data sovereignty dimension: locally-run agents are particularly relevant for healthcare given the low energy prices in the region and the sensitivity of medical data, with low latency and data localisation both practical operational advantages rather than theoretical ones.

Smart-city operations are another early fit, where physical AI can layer onto existing sensor networks, asset management systems and automation stacks already being built across Gulf projects. This blend of greenfield environment and state-backed procurement appetite is rare globally – most markets require physical AI to prove itself against existing infrastructure. In much of the Gulf, the infrastructure is being designed with these systems in mind.

��

The Future-Focused Playbook for Founders and Operators

��

Morales challenges the dominant belief that physical AI is strictly a hardware phenomenon. “What founders and operators need to think about now is integration, not innovation,” he says. “The technology stack is now largely commoditised. The competitive advantage in the next 24 months will belong to whoever figures out how to deploy these tools inside existing operational structures without requiring a full rebuild. That means deep partnerships with regional operators, not just pitching to government programmes.”

The winners in this field will likely be companies that can localise hardware support, data pipelines, safety systems, maintenance and operator training for regional conditions – the heat, dust and fast-changing infrastructure that characterise Gulf deployment environments. Proving ROI in narrow, high-frequency use cases first, rather than pitching all-encompassing robotic solutions, is the approach most consistent with how large Gulf operators actually make procurement decisions.

Morales puts the broader opportunity straightforwardly: “ is not just a deployment zone. It’s one of the few places on earth where the capital, the political will, and the infrastructure ambition exist at the same time. That window is open now.” The Gulf may not invent every robot first. But right now, it’s building the environments where those robots will learn to operate at scale – and that might matter more.

The post Why The Gulf Could Become The World’s Most Exciting Deployment Zone For Physical AI appeared first on 91̽��.

Mythos was kept within Project Glasswing, a restricted defensive programme. Access was granted to a limited group of organisations including Amazon, Microsoft and Apple for cybersecurity work.

Anthropic believed defenders should receive advanced capabilities before attackers gained access to similar systems at scale. That policy made Mythos one of the most talked about AI systems in cybersecurity circles.

The upcoming release changes access from a select group of organisations to the company’s entire customer base. Cybersecurity professionals and AI companies are considering what this could mean once availability expands.

Reuters reported that Mythos can identify software weaknesses, evaluate exploitation possibilities and reason through attack paths that once required elite cybersecurity expertise. Those capabilities explain why the model has generated so much interest.

��

Why Has The Release Generated Conversation Among Experts?

��

Security specialists have spent months discussing the consequences of making advanced vulnerability research capabilities available to a much bigger group of users.

According to Reuters, restricted access because it believed defenders could repair dangerous vulnerabilities before attackers obtained comparable technology. That philosophy guided the company’s handling of Mythos through Project Glasswing.

The upcoming release changes that arrangement. Many conversations now examine what happens when advanced cybersecurity capabilities become available to a much larger population.

Gil Geron, CEO of Orca Security, said the issue extends far beyond one product release.

“The security concerns surrounding Anthropic’s plans to publicly release Mythos aren’t really about this one release. They reflect a bigger shift. AI is no longer confined to models sitting in isolation. It is embedded directly into business operations, decision-making, and automation. That is where the real risk begins to scale.”

His comments place emphasis on how advanced AI systems interact with business environments, company data and automated workflows.

��

How Are Security Companies Reacting?

��

Many cybersecurity companies are talking about operational exposure and system access more than the model itself.

Geron said the biggest issues arise when AI systems gain access to infrastructure, company data and automated processes. He said, “As AI capabilities advance, the advantage does not stay evenly distributed. Attackers are quicker to operationalise these tools and are not slowed down by governance, internal controls, or risk tolerance. That creates asymmetry where innovation moves fast, but defence lags behind. The gap is not only growing, it’s accelerating.

“What organisations often miss is that model safety is only one layer of the problem. Risk is introduced through access, integration, and execution. Which systems the AI can touch, what data it can pull in, and how confidently its outputs are acted on. These are the real exposure points.”

That reaction puts emphasis on access permissions, and automated actions. Many cybersecurity teams view those areas as important sources of exposure.

Geron also said organisations need much deeper visibility into how AI systems operate across their environments.

“Without continuous visibility into how AI is actually being used across the environment, that is a risk multiplier, not a safeguard.”

Those comments explain why a lot of the reaction extends past just software vulnerability research.

��

Has Anthropic Changed The Model Before Release?

��

Not everyone expects the version released to customers to match the version used within the restricted programme.

Mohammad Moahid, Co-Founder and Managing Partner of Zero To Agent, believes the public version may contain limitations added before release.

“Even though Mythos is widely reported to have been an internal name for the Opus 4.8 model. This has been fine-tuned and is no longer finding zero-day exploits amongst every single legacy architecture. On top of that Anthropic has gifted its selected organizations with the head start to patch any outstanding vulnerabilities. We believe that the released mythos model has been nerfed in its obedience to certain tasks that may be flagged as dangerous.”

His assessment is based on the possibility that Anthropic adjusted the model’s behaviour before making it available to all customers.

Selected organisations received time to patch vulnerabilities before broader access became available. That sequence formed an important element of Anthropic’s handling of Mythos.

The release will provide the first opportunity to see how those restrictions perform when Mythos reaches a much bigger audience. Anthropic’s experiment with controlled access is entering a very different chapter.

��

More Experts React To Anthropic’s Plans

��

AI experts have also reacted to the news…

��

Our Experts:

��

• Olli Krebs, SVP EMEA, Incode
• Dr. Dominik Hörndlein, AI Strategy and Implementation Consultant, Hoerndlein Consulting
• Stephanie Herder, Executive Business Growth and Senior Lead Project & Process Management, Specific Group
• Promise Akwaowo, Process Automation Analyst, Royal Mail Group
• Andrellos Mitchell, BSW, MA, JD,Attorney, Legal And Policy Analyst, Publisher, The Mitchell Report
• Jonathan Beresford, Founder, MathsTutor
• Ben Rometsch, Co-Founder and Chief Technology Officer, Hoxton Mix
• Sheraz Ali, Founder, HARO Links Builder
• David Moosmann, Founder, LearnClash

��

Olli Krebs, SVP EMEA, Incode

��

��

“Anthropic publicly releasing Mythos is significant because it shows that frontier AI cybersecurity is moving from being available to a small number of select companies to a capability that could reshape how the entire industry approaches cybersecurity.

“The positive story is that AI is becoming exceptionally good at understanding code, identifying patterns, and finding vulnerabilities at a scale humans simply cannot match. That gives defenders the potential to find and fix security issues much faster than before.

“But the bigger story is that these capabilities work both ways. The same technology that helps security teams discover weaknesses can also accelerate reconnaissance, vulnerability discovery, and the adaptation of new attack methods. As these tools become more widely available, the speed of both attack and defence is likely to increase.

“This is an important moment because it highlights that cybersecurity can no longer be treated as a static problem. Organisations will need security systems that continuously evolve alongside AI-driven threats. It also reinforces why identity and trust are becoming increasingly important. As AI becomes more capable of acting on behalf of people, the industry will need stronger ways to verify who is real, who is authorised, and when a machine is operating without a human behind it.

“Mythos is not just another AI release. It is a signal of where cybersecurity is heading next.”

��

Dr. Dominik Hörndlein, AI Strategy and Implementation Consultant, Hoerndlein Consulting

��

��

“The wider release of Mythos will accelerate the patching of long-neglected vulnerabilities. But a powerful model alone won’t reshape the security landscape.

Mythos and the competitive landscape

Anthropic’s decision to move Mythos beyond Project Glasswing and into wider availability is a significant step – but it is worth tempering expectations. Anthropic is good in building anticipation, and Mythos is a perfect example where we see this ability play out. As an example, the company maximises press coverage through restricted access to create a fear-of-missing-out, followed by a timed public release to profit from the anticipation they have built up.

The model seems to be genuinely capable, but the idea that it represents a qualitative leap far beyond what competitors offer is almost certainly overstated. What Anthropic has is a brand narrative, not an unassailable technical moat.

What the release will actually change

“Where Mythos will make a real difference is in the short-to-medium term clean-up of historic technical debt. Large parts of the internet run on open-source software maintained by one or two developers in their spare time. This is the reason why we face plenty of code with vulnerabilities – unpatched for years, sometimes decades.

“A model capable of autonomously scanning, hypothesising, and testing attack paths at machine speed will surface many of these bugs quickly. That is a valuable contribution, and the security community should welcome it.

“In the longer term, the speed of this iteration will force the industry to build patching and update pipelines that are themselves increasingly automated – the bottleneck will shift from finding vulnerabilities to closing them fast enough.”

The part the headlines always miss

“But a powerful model is only one component of an effective security solution. The practical gain comes overwhelmingly from how that model is integrated into a broader ecosystem of tools: automated patching pipelines, zero-trust network architectures, dependency scanners, and rapid update cycles.

“Without that surrounding infrastructure, even the best model produces findings that sit in a backlog. The true game-changer is not itself, but AI models like Mythos embedded in that broader ecosystem – because computers are simply much faster than any human IT security expert.”
��

Stephanie Herder, Executive Business Growth and Senior Lead Project & Process Management, Specific Group

��

��

“Mythos should not be made widely available simply because it may help defenders. A system that can find digital ‘unlocked doors’ in minutes could expose hospitals and public services before fixes are ready. Attacks move at machine speed; repairs still take people, time and money.

“Used defensively under strict supervision, Mythos may help close dangerous gaps. What makes this harder to accept is that the public is being asked to trust controls it cannot yet properly assess. Anthropic has described supervised access, but too little is publicly known about how that supervision is enforced in practice. Reports of unauthorised access through a third-party environment only deepen the concern that AI capability is advancing faster than public accountability.”

��

Promise Akwaowo, Process Automation Analyst, Royal Mail Group

��

��

“Anthropic’s planned public release of Mythos is significant because it moves AI capability from general productivity into a much more sensitive area: autonomous or semi-autonomous cybersecurity work.

“From my perspective working across enterprise automation, AI-enabled delivery and governance-led digital transformation, the biggest issue is not simply whether Mythos can find vulnerabilities. The real question is whether organisations are ready to govern a tool that can accelerate both defensive and potentially harmful cyber activity.

“For the industry, Mythos could be extremely valuable if used responsibly. It could help security teams identify weaknesses faster, support overstretched engineering teams, and improve resilience across critical systems. However, a wider release also increases the risk that powerful technical capability becomes accessible to people or organisations without the right controls, maturity or accountability.

“This is where responsible AI governance becomes essential. Access to tools like Mythos should not be treated like access to a normal chatbot. Organisations will need clear usage policies, role-based access, audit trails, human approval points, monitoring, and escalation routes when the model identifies high-risk vulnerabilities. The human-in-the-loop should not be symbolic; it must be part of the operating model.

“In general, this release shows that AI governance can no longer sit only with data science or legal teams. Business analysts, product leaders, security teams and executives will need to work together to define where automation adds value, where it creates risk, and who remains accountable when AI systems act at speed.

“My view is that Mythos represents a turning point: the industry must move from asking “what can AI do?” to “what should AI be allowed to do, under whose supervision, and with what evidence of control?””

��

Andrellos Mitchell, BSW, MA, JD,Attorney, Legal And Policy Analyst, Publisher, The Mitchell Report

��

��

“Anthropic’s plans to publicly are significant not because of the technology itself, but because of what it says about the future relationship between artificial intelligence and society.

“The AI industry has spent years promising that more powerful models will improve productivity, creativity, and innovation. Those benefits may well occur. However, every new generation of AI also raises new questions about trust, accountability, employment, education, misinformation, and public understanding.

“For the average person, the release of Mythos likely will not be remembered because of its technical specifications. It will be remembered as another step toward a world where AI becomes more deeply integrated into daily life, business, government, and decision-making.

“The larger issue is that technological development continues to move faster than public discussion about its consequences. As companies race to release increasingly advanced systems, policymakers, educators, employers, and ordinary citizens are struggling to keep pace.”

��

Jonathan Beresford, Founder, MathsTutor

��

��

“My view is that publicly releasing a system like Mythos would be a shift from frontier models as private lab capability to frontier models as ecosystem infrastructure. The upside is faster independent testing, more external research, and more practical innovation from teams who cannot train models at that scale themselves. The risk is that release changes the threat model: once a powerful model is broadly accessible, misuse is no longer theoretical and safety work has to move from controlled evaluation into monitoring, access design, rate limits, auditability, and clear downstream accountability.

“For the industry, the important question is not simply whether Mythos is powerful. It is whether Anthropic can make the release boring in the operational sense: documented limits, clear safety boundaries, strong abuse detection, and enough transparency for serious users to understand where the model should not be trusted.

“In education and edtech, this matters because powerful AI systems are tempting to treat as universal tutors. They are not. A good learning product still needs curriculum constraints, age-appropriate scaffolding, worked reasoning, and checks against confident wrong answers. Publicly available frontier models can accelerate useful tools, but only if builders wrap them in narrow, testable product design rather than letting the model become the product.”

��

Ben Rometsch, Co-Founder and Chief Technology Officer, Hoxton Mix

��

��

“As a CTO and software engineer, it’s almost impossible to overstate how much professional life changed towards the end of 2025.”

“Up until 2026 we’d been writing code, line-by-line (by hand!) for over twenty-five years. In the last six months, I haven’t written one line. What happened? Anthropic, the makers of Claude, have a tool called “Claude Code” that helps write code for you with the aid of their LLM models. On November 24th, Anthropic release the Opus 4.5 model. Prior to Opus 4.5, the backing models were weak, struggling to understand larger projects and more complex logic. Opus 4.5 changed everything; overnight, you could ask Claude Code to plan and execute large features, and the code it generated was *really* good.”

“Since then they have released Opus 4.6, 4.7 and just a few days ago 4.8. Each of these versions of Opus have gotten incrementally better. They are now so good we are running out of engineering work and asking our product team for more features. This never happened!”

“Then Anthropic announced Mythos.”

“Mythos looks to be an even bigger leap than Opus 4.5 was. Given the revolution that we engineers experienced with the introduction of Opus, this announcement was absolutely huge. But there was a problem; a big problem.”

“Mythos was so good at understanding code and the underlying systems that they run on, that Anthropic found hundreds of security vulnerabilities in large pieces of software infrastructure. If they had released Mythos to the general public at the time, it would have most likely taken down the entire internet. That’s not hyperbole. Core engineers of the Linux operations system (which powers most of the Internet) found dozens of security issues using Mythos, putting businesses at risk after being given private access to the tool.”

“The mind-bending thing for engineers is that Opus has already changed our world. I will most likely never write more than a few hundred lines of code by hand for the rest of my life. So to hear that, within six months, they are announcing something that is another step change, is beyond belief.”

��

Sheraz Ali, Founder, HARO Links Builder

��

��

“While the world rejoices at Anthropic releasing their AI to detect software vulnerabilities in over 23,000 different applications, as a digital agency CEO, I see it through a more skeptical lens. This type of vulnerability scanner doesn’t benefit the good guys alone – it grants the same initial advantage. With the half of the internet consisting of open-source software and running on WordPress, security was never great, and both parties have access to the same technology while one is generally quicker than another.

“Keeping their AI, named Mythos, under lock and key until now wasn’t an accidental choice made by Anthropic in their project called Glasswing. As reported, this machine is capable of writing cyber attacks in addition to finding vulnerabilities within software. The moment when this happens, the clock starts ticking, and small businesses that lack their own cybersecurity teams won’t be able to compete in that race. Already, Mozilla released an update of Firefox containing patches for 271 vulnerabilities detected by Mythos.

“Bottom line: security is shifting from an IT concern to a matter of survival. Companies which respond to the public release of Mythos and promptly perform a vulnerability scan will seem savvy after six months. On the other hand, companies that ignore this information and get hacked in the meantime will probably be busy writing apology letters.”

��

David Moosmann, Founder, LearnClash

��

��

“The honest reaction from my desk: each Anthropic jump compresses my dev cycle in a way that’s hard to describe unless you ship every day. Opus 4 last year, about 5 retries per feature, and the code still limped after. Opus 4.7, mostly one-shots, maybe a cleanup pass. So if Mythos really sits another tier above 4.7 the way Anthropic is saying, the slow part of my day stops being “wait for the model to get it right” and starts being “did I make the right architecture call?”. That’s a different bottleneck. A more interesting one, honestly.

“The security headline (Linux-kernel exploit chaining, 73% on expert tasks) is properly scary and will obviously run with most of the coverage. What I think matters more for , though, is what wide Mythos access does to solo shipping. For texture on what solo shipping at this level looks like: LearnClash is 442 Dart files plus 168 TypeScript Cloud Functions, four languages, 17 feature modules. The pre-AI version of me would’ve needed a five-person team for 12-18 months.”

The post AI Experts React To Anthropic’s Plans To Publicly Release Mythos appeared first on 91̽��.

87% of UK business leaders expect quantum computing to disrupt their sector by 2030. This is according to EY’s Quantum Business Readiness Report, produced with the National Quantum Computing Centre and after surveying 500 leaders from companies generating more than £150m in revenue, this is what they found:

That expectation has already turned into planning activity, with EY finding that 35% of UK business leaders have made quantum computing a strategic priority during the next five years.

��

Which Industries Are The Most Interested So Far?

��

Financial services recorded the highest level of interest, where 67% have given quantum computing that status, whereas real estate, hospitality and construction recorded a much lower figure of 17%.

The sectors involved are uses that could deliver commercial value. Banks and payment processors are assessing quantum computing for fraud detection and anti money laundering work, while automotive manufacturers are assessing how it could assist electric vehicle battery development and traffic management.

What makes the findings interesting is that executives seem convinced quantum computing will have a substantial impact on business operations, even though many do not expect that impact to arrive immediately.

��

If Executives Expect Disruption, Why Are They Not In A Rush?

��

The EY research brought up an interesting contradiction where almost 9 in 10 business leaders expect quantum computing to disrupt their sector by 2030, but 59% believe the technology will not mature enough to become important in core operations until 2030 or later.

That view can also be seen in hiring plans, with only 13% of UK business leaders intending to recruit the quantum talent they expect to need during the next two years.

But also, 83% believe losing competitive advantage is associated with failing to adopt the technology, creating an unusual gap between long term expectations and immediate action.

Piers Clinton Tarestad, Technology Risk Partner at EY UK, said, “Our latest report shows that, while UK businesses are continuing to invest in and evaluate the impact of quantum computing, expectations of its maturity remain cautious. This means that although long-term disruption from quantum computing is widely expected, including its impact on cybersecurity, the proportion of companies taking immediate steps to prepare remains relatively small.”

He also explained why many executives struggle to get started. “For some business leaders we speak to, hesitation stems from not knowing where to begin exploration. One effective first step is to identify one or two practical ways in which quantum can be used, linked to existing pain points, before assessing which teams or processes would be most affected, and any gaps in skills, tools or data.

“In the same way that an orchestra needs a conductor, effective quantum implementation needs a senior sponsor within a business with cross-functional oversight to coordinate efforts, secure resources, and help ensure quantum readiness becomes part of broader digital transformation plans.”

Early shares quite a few of these similarities, where many organisations recognised its long term value long before they committed substantial resources to it.

��

��

Could Cybersecurity Become Quantum’s First Big Battleground?

��

Many discussions about quantum computing revolve around scientific breakthroughs, but security specialists often view the technology through the risks it could create for existing systems.

EY found that 81% of business leaders see the outdating of existing IT systems as a business risk associated with quantum computing, while 80% identified future regulatory compliance as a risk.

David Holtzman, Chairman of Naoris Protocol, believes cybersecurity may become the area where quantum computing has its earliest commercial impact.

He said, “Quantum computing is absolutely a candidate for the next major technology wave after AI, but its impact will be felt most acutely . If I were a CISO or CIO today, the first thing I’d do is audit where encryption is being used across the organisation. Most people would be shocked by how much of their business depends on cryptography.

“The same is true for AI. Most AI systems rely heavily on encryption to protect data, models, communications, and identities. If that encryption breaks in a quantum era, much of the trust underpinning those systems breaks with it. In many cases, if the encryption falls, so does the autonomous or agentic device relying on it.”

That perspective helps explain why quantum computing receives so much interest long before large scale commercial deployment becomes commonplace, particularly when security infrastructure can take years to replace.

��

Is Quantum Really The Next Wave After AI?

��

Perhaps the question itself misses what could happen.

Author Pranav Bhatnagar believes AI and quantum computing will develop together, instead of one replacing the other.

He said, “I don’t believe quantum computing is after AI. I think the two technologies will grow together, with AI becoming the bridge that helps bring quantum computing into practical use.”

Bhatnagar used electricity as an example, explaining that the technology existed long before smartphones transformed everyday life, but its biggest influence came when other technologies were built on top of it.

He believes quantum computing could follow a very similar path by finding commercial value through drug discovery, materials science, climate modelling, logistics and optimisation.

He also spoke about and warned on the timeline, saying, “That said, I think we’re still much earlier in the quantum journey than many headlines suggest. AI is already affecting businesses, jobs, education, healthcare, and cybersecurity today. Quantum computing still faces significant challenges around scalability, stability, cost, and practical deployment.”

His final observation may explain why business leaders are paying attention even though commercial use remains relatively low. He said, “The real opportunity is not asking whether quantum will replace AI. The bigger question is what becomes possible when AI and quantum computing eventually work together. That combination could unlock breakthroughs that are difficult to imagine using today’s technology alone.”

That may prove to be the most realistic outcome. AI has already entered everyday business life, whereas quantum computing is currently a technology that organisations are studying and/or budgeting for. The next decade may show that the was never AI versus quantum, but AI and quantum computing developing together.

The post Is Quantum Computing The Next Wave After AI? Experts Share Their Insights appeared first on 91̽��.

That picture is changing – MBZUAI in Abu Dhabi, KAUST in Saudi Arabia and a wave of AI-focused academic programmes across the region are now producing graduates specifically trained for the AI economy. Saudi Arabia’s SAMAI programme trained 1.2 million participants in under a year, reaching its three-year target ahead of schedule. Qatar has certified 13,000 professionals in AI skills as of 2025, targeting 50,000 by 2030. MBZUAI’s retention data is notable too: nearly 80% of alumni continue working in the UAE’s AI sector within their first year

That said, the fundamental problem is still unresolved. The UAE holds just 0.7% of global AI talent, ranked 16th worldwide. Saudi Arabia holds 0.4%, ranked 19th. Despite salary packages exceeding $1 million, the Gulf still struggles to attract senior AI researchers at scale, because mature Western hubs offer network density, equity upside at high-growth startups and peer research communities that took decades to build. The pipeline is real, and the sustainability question is still open.

��

Education Alone Has Never Been Enough

��

The retention challenge is the crux of it. Producing graduates is a solvable problem – every government with money and intent can build a university. Keeping those graduates once they’re good enough to have options is harder.

The data from MBZUAI suggests the Gulf’s approach to this is working better than expected, with long-term residency pathways, competitive compensation and a clear narrative around building AI for regional challenges rather than serving global platforms all contributing to higher-than-expected retention rates.

Education in the region is growing faster than the commercial AI sector around it, which creates pressure on the most talented graduates to seek environments where they can work on consequential problems immediately after graduating. The Gulf’s solution to that challenge is building both simultaneously – universities and the AI startup infrastructure for graduates to work in. Whether that commercial layer can develop quickly enough is the central question.

We asked five experts across AI education, talent development and Gulf business to share their take on the matter.

��

��

Our Experts

��

• Corina Goetz: Gulf Business Culture Specialist, Star-CaT
• Pranav Bhatnagar: Book Author and AI Commentator
• Mariusz Bajorek: Founder and Strategic Lead, Momentum Bridge Consulting
• Edward Tian: Co-Founder, GPTZero
• Jitesh Keswani: CEO and Founder, e intelligence

��

Corina Goetz, Gulf Business Culture Specialist, Star-CaT

��

��

“I can only answer for the Gulf. MENA is a category that flattens very different countries, and I will not pretend to speak for places I do not work in.

“Inside the Gulf, the question itself is the problem. or homegrown talent – pick one. The Gulf is doing both, deliberately, at the same time, and the West keeps missing it.

“MBZUAI in Abu Dhabi is training graduates on regional problems, not Bay Area problems. KAUST is doing the same in Saudi. Real institutions, real research output. But the more interesting move is what is happening around them. UAE Golden Visas for AI specialists. Equity stakes at G42. Housing packages at NEOM that would embarrass a Series B startup in San Francisco. Saudi salaries that London cannot match. The Gulf is not waiting for its graduates to be ready in ten years. It is importing senior talent now while training the next generation, and it is paying enough to keep both.

“Is the pipeline at the scale the ambition requires? Not yet. Will the region depend on imported expertise for another decade? Yes. But I have sat in offices in Abu Dhabi and Riyadh where the AI teams are half Emirati or Saudi, half hired from London, Cambridge, MIT. That is not failure. That is how Silicon Valley was built in the 1970s. The Gulf understands this. The West has not noticed.”

��

Pranav Bhatnagar, Book Author and AI Commentator

��

��

“MENA is at a very interesting turning point. A few years ago, the conversation was mostly about attracting AI talent from outside the region. Today, it is increasingly about whether the region can develop and retain its own talent at scale. The fact that institutions like MBZUAI, KAUST and other AI-focused programmes exist at all is already a sign of progress, because they are creating a generation of graduates specifically trained for the AI economy rather than adapting traditional pathways later.

“That said, education alone does not create a sustainable talent pipeline. People stay where they find opportunity, mentorship, challenging work, funding, and the ability to build meaningful careers. If the most ambitious graduates still feel they must move to Silicon Valley or London to access world-class research, startup ecosystems, or career growth, talent leakage will continue regardless of how many graduates are produced.

“The encouraging sign is that MENA is no longer investing only in education. The region is also investing in , research, startups, sovereign AI initiatives, and innovation ecosystems. That combination is far more powerful than education alone. The real measure of success over the next decade will not be how many AI graduates MENA produces. It will be how many choose to build companies, conduct research, create intellectual property, and lead AI innovation from within the region itself.”

��

Mariusz Bajorek, Founder and Strategic Lead, Momentum Bridge Consulting

��

��

“The honest answer is: both things are true at the same time, and that’s what makes this question so difficult. MENA is genuinely building something. MBZUAI and KAUST are producing graduates who can compete globally – I’ve seen CVs coming out of Abu Dhabi that would get interviews at any European or US tech company. That was not the case five years ago.

“But “can compete globally” is exactly the problem. The moment a graduate is good enough to work in San Francisco or London, they have a real choice. And right now, compensation, visa pathways, and the density of high-growth companies still tip the scales westward for many of the best ones. What I’m watching from the talent market side is whether MENA can close that gap fast enough – not just with salaries, but with the kind of career trajectory that makes staying feel like the ambitious choice, not the safe one. That shift is starting. But it’s not there yet.”

��

Edward Tian, Co-Founder, GPTZero

��

��

“MENA is creating real opportunities to train people on how to use AI. Universities like MBZUAI and KAUST are creating technically competent graduates. The problem is after they graduate. The best students are looking for environments that have dense research activity, rapid product development, and strong peer relationships. Those ecosystems still exist primarily in Silicon Valley, London, and a few other places.

“I see a timing issue here. Education is growing faster than the industry can grow. This creates an outward pull on people training in the local areas. Retention is better when graduates can work on production systems soon after they graduate. The sooner they can work on projects with real users and get feedback loops, the better. The pipeline is real. The sustainability question hinges on whether the local ecosystems can develop as fast and as deeply as the best of the global AI world.”

��

Jitesh Keswani, CEO and Founder, e intelligence

��

��

“The question is not whether the MENA region can produce people who are good at AI. The region is putting money into universities, research and government programmes. The big question is whether it can give these people jobs so they stay for a long time.

“What is encouraging about MENA now is that people are talking about building a community for AI, not just teaching it. We see investment going into AI startups, companies using digital technology, and countries making plans for AI that create real jobs for skilled people. Keeping these people will depend on whether they can have good careers where they live. If the best projects, the fastest-growing companies and the startup capital are still elsewhere, people will still leave.

“MENA is starting to have its own people who are good at AI. But we should not say it is successful just because many students finish AI programmes. We should say it is successful when these people stay, start companies, come up with new ideas and create more AI opportunities in the region itself. The next five years will be about giving people reasons to stay, not just making people who are good at AI.”

��

For any questions, comments or features, .

��

The post MENA Is Building AI Universities – Is A Homegrown Talent Pipeline Finally Here? appeared first on 91̽��.