AI is transforming the way MENA financial institutions loan to people and small businesses

1

The Pitch

If PHP2Shell woke you up to the importance of cybersecurity, this guide is for you

Good morning, wonderful people. It’s an insane news morning after Anthropic finally filed for its IPO and Meta said it was looking to raise USD 95 bn in a series of rolling offerings, all while OpenAI is in the market with another big private raise for which it’s hoping to attract fresh capital from the UAE.

Oh, and Microsoft is offering buyouts to quarter of its global workforce, Oman is making a data center play, and Humain may have landed its first big offshore client.

We also have for you a deep dive into how AI might change credit decisions for personal and small business loans.

BUT FIRST- Sometimes, the only sane thing to do is to pull the plug. That’s the lesson of the last week or so as big organizations including the parent of Wikipedia, units of telecom giant Orange, European government agencies, and the Internet Archive took their websites offline as they scrambled to figure out how to patch a new security flaw.

Why go dark? The worry was that AI would help attackers exploit the vulnerability before the teams running the sites could get the fixes in.

AI nerds on X have taken to calling it Black October, saying the world is at a tipping point where attackers with AI can now overwhelm those whose job it is to keep our data safe and sites and apps online. The commentary has ranged from bog-standard advice about updating your systems to suggestions that you stockpile cash and food.

At the center of last week’s scare was something called PHP2Shell — flaws in PHP, a type of code that experts say is core infrastructure for something like 70% of the web, including WordPress and plenty of other websites and business applications. The exploits could let an attacker steal information or run their own code on an affected server. The PHP project released emergency fixes on 21 October, and it looks like they got there in time: I couldn’t find a single report of a site in MENA (or anywhere else) going down to a third-party attack.

If you’re like most businesses, your biggest vulnerability is probably sitting in a broom closet or a little-visited corner of your server room. Your head of IT knows which one, but if you’re poking your head in by yourself, it probably looks something like a rectangular server box with a sign on it that says something to the effect of “Important! Don’t shut down!” What does it do? What happens if you power it down? Who built it? Is s/he still alive? Who knows. (Tap or click here for a spoof so funny only because it’s so real and while you’re there, cue the Twitterbros.)

The bottom line: Someone needs to figure out why the server / service / whatever exists before you can make a decision to patch, rebuild, or switch it off. And you want to do that now, while you have the luxury of time, not after an attacker locks up your customer’s data.

AI is making it much easier for attackers to figure out how to get into your website and other systems. Here at EnterpriseAM, Joe has reported on how AI is also making factory machinery and power grids easier to hack. Sherif Shaltout (LinkedIn), managing executive for MENA at Liquid C2, told me earlier this week that vulnerabilities once considered theoretical or too hard to exploit now face “agents and tools that are relentless and have all the time in the world to figure it out.”

Your tech team may need some extra time to get its act together. Mario Tomić, our chief audience and technology officer, says there are literally thousands upon thousands of PHP applications still in service across our region still doing the one thing they were built to do years and years after they came into service. The operating system may no longer be supported, and updating PHP may require replacing the whole damn server. The Catch-22 is that if you do that, the old application might just … stop working.

“So you need to kind of throw away the whole thing and rebuild it,” Mario says. “Often, the old PHP code won’t work. For example, PHP 5 won’t work on PHP 8.”

That’s the challenge beyond this week’s patches. AI is speeding up the pace at which attacks are coming — and how quickly they’re executed — faster than most companies can identify weaknesses, figure out how to patch or replace systems, get the budget they need for the project, and deploy teams.

But waiting for the next emergency to show you what needs rebuilding could be an expensive bloody bet. Here’s what Sherif and Mario say you need to do next:

#1- Start with a map. Your team needs to know what software every part of your business runs on, where it lives, what it connects to, and who’s in charge of maintaining it, Mario says. “Don’t assume the map exists,” he adds. The map probably includes stuff you’ve bought and now integrate with as well as applications somebody built years ago. Cast a wide net: Shaltout describes companies that think they don’t use PHP because they use WordPress — which itself runs on PHP. He urges businesses to retire forgotten applications that no longer serve a purpose, focusing resources on maintaining only those things they really need and use.

#2- Give your CISO a proper budget. Pay for upgrades, replacements, and testing before an urgent fix forces you to choose between an exposed system and downtime for a critical part of your business. “It’s very hard to upgrade,” Tomic says of those older installations. Postponing the job won’t make the next emergency easier.

Uhm, Enterprise? You mean my Cisco? (IYKYK). No, I mean your chief information security officer — the person responsible for protecting your systems and data. If you’re asking and your business brings in more than a few tens of mns of AED a year, odds are good you’re underinvesting in cybersecurity. Read more about the role here.

And put more people on the problem. Shaltout’s warning is that AI needs experienced people directing it. “I think we need to find the balance between how quickly we fire people” — he laughs — “and how do we get them to guide AI to write that better software because it’s not going to do it on its own. You need to have the right people, and they need to have the right tools to defend.”

#3- No more box-ticking: Shaltout points out that too many businesses (particularly smaller institutions in regulated industries) have approached cybersecurity from a compliance point of view, meaning that, “you risk it becoming checkboxes and checklists and not fundamental questioning of systems and processes. It has to come from a risk-management point of view, not fear of a fine or a penalty from a regulator.”

Saudi businesses tend to be among the best in the region on cybersecurity, Shaltout adds, saying the National Cybersecurity Authority ensures that “everyone needs to submit a level of how compliant they are, regardless of their size, their industry.”

#4- Go hack yourself. Mario argues you run penetration tests against your own systems with capable open-source models. In practice, that usually means open-source and open-weight Chinese models, because frontier models from Anthropic and OpenAI generally refuse to look for vulnerabilities unless you’re an approved security researcher — they don’t know the difference between you testing your own systems and a hacker going after you. The idea is to find and fix weaknesses in your own systems before somebody else does, he says.

You don’t have to buy 100% of the “Black October” hype to act. Ask your team to figure out what that server does. Put somebody in charge and then pay for the upgrades. That sign on the door warning you not to turn off the box? It’s not a maintenance plan. –Patrick

Can AI help make lending fairer in Egypt and the Gulf?

AI is transforming the way MENA financial institutions decide who’s a good loan prospect, stoking excitement about a more inclusive market, but also anxiety about who might get left out.

On the plus side: AI models are making it easier to assess the creditworthiness of small businesses and individual borrowers, especially newcomers to the financial system and lower-income people without bank accounts. These people often have little or no conventional credit history, but an AI model can use non-conventional data to make a decision. For people wanting personal finance, that can mean combing through their phones to look at how they pay their phone and utility bills, Ahmed Abaza (LinkedIn), co-founder and CEO of Synapse Analytics, tells EnterpriseAM.

It’s largely about cost. Making a decision on every personal or small business loan by hand takes so much staff time that it can eat up much of what the lender earns on it, the World Bank has noted. Letting software make those decisions can cut that cost significantly, McKinsey says, making those segments — as well as microfinance — more attractive to banks and non-bank lenders alike.

But there’s also a less rosy version of the story: AI is trained on existing data — and existing biases. If women, informal workers, and young people barely show up among a lender’s past borrowers, a model can end up shutting them out.

The developments highlight how the shift to AI can dramatically increase efficiency while simultaneously removing the human elements of any business transaction — a tradeoff that’s especially troubling when it comes to personal, home and business loans, which can be life changing for many borrowers.

The balancing act: “I use [AI] to increase human efficiency, not to replace people, because a human must guide it and assess the output, as AI can often be inaccurate,” Hossam Heiba, managing director of Orbis Capital and former head of Egypt’s state investment agency, GAFI, tells EnterpriseAM.

Big corporate loans are a whole other ballgame. Some banks are already using AI to help judge whether a company can handle a loan, Bassem Fayek (LinkedIn), managing director and partner at BCG, tells EnterpriseAM. But handing the final call to a machine is a much harder sell when a bank’s reputation is on the line. “I answer to the depositors and the shareholders. But the buck stops with me and my team,” Todd Wilcox, CEO and Deputy Chair of HSBC Egypt (LinkedIn), said at the EnterpriseAM Egypt Forum earlier this month. “It’d be difficult to defend to depositors that the bank failed because of our AI model and [its] judgment.”

Regulators in Egypt are starting to ease the path for more machine decisionmaking in lending. From April 2027, consumer-finance companies deciding whether to approve financing will have to consult a gauge from the country’s only licensed private credit bureau based on behavioural analysis of alternative data, per a new rule from the country’s non-bank financial regulator. The score is the type of information that feeds directly into AI-backed risk prediction and automated lending decisions.

Why it matters: Fewer than half of Egyptian adults have a bank or mobile money account, according to the World Bank, which makes it extremely difficult to assess their creditworthiness using traditional data and approaches.

Credit scoring is even more advanced in the UAE. Since July, credit scorers have begun incorporating how people repay micro-loans from buy-now-pay-later firms such as Tabby and Tamara in their credit reports.

And they’re increasingly interested in assessing those reports using AI agents rather than humans alone. Emirates NBD is set to become the first MENA bank to run ChatGPT on infrastructure inside the UAE. The bank recently ranked first among 25 banks in the Middle East and Africa edition of the Evident AI Index, which ranks banks on how far along they are in adopting AI. Emirates NBD declined to comment for this story.

Why that matters: UAE banks aren’t allowed to send customers’ private data out of the country without the central bank’s approval, and ChatGPT normally runs in data centers abroad. With ChatGPT running inside the UAE, the data never leaves the country, so the bank can use it on its own data.

Advocates say there’s an open-and-shut case for turning lending decisions over to AI agents — they’re smarter, faster and more capable than humans. Think of it as hiring “the best risk agent you can ever hire, that never sleeps, never eats, never takes time off,” who’s always “exploring how to have you accept more customers with less risk,” Synapse’s Abaza tells EnterpriseAM.

The numbers: Synapse, whose software is used by more than 50 financial institutions, says its clients are approving up to five times more customers with 40% fewer bad loans compared to before they signed up with Synapse. Abaza says his clients are not lowering the bar — they’re just getting through far more applications. Ultimately, banks that once approved around 10k people a month may end up approving 10k a day, Abaza says.

MNT-Halan, another Egyptian fintech, is using its in-house tool, Apollo, to tell staff which overdue loans to chase each month, as we previously reported. It’s set to roll the tool out across “many departments” within three months, CEO Mounir Nakhla said at our Egypt Forum, just weeks before the company’s Egyptian unit debuted on the EGX.

Flying solo: The Egyptian fintech Klivvr even expects to soon let AI agents make decisions about underwriting loans without any human approval, as we reported last week.

But there’s also anxiety about moving too fast to AI, especially at more traditional banks. Executives there worry about giving powers to AI without ensuring that its decisions are properly vetted and that there’s a human taking responsibility. “If you have poor governance and no accountability, it simply leads to making bad decisions faster,” HSBC Egypt CEO Todd Wilcox said at our Egypt Forum.

People, not algorithms: Even Emirates NBD, which styles itself a regional leader in the adoption of AI, insists in its published AI policy that “people, not algorithms, remain in charge.”

Important caveats: AI is only as good as the data it’s fed, and most companies’ data is a mess. That’s true for big firms and small ones, in every industry and almost everywhere, Hazem Gebili, chief AI and digital officer at Orange Egypt, tells us. Records tend to sit in old systems scattered across a company, much of it inaccessible to AI. And just plugging AI into the way things already work would be “the biggest failure,” he says: “You need to reimagine the journey.”

Most banks are keeping AI away from their core infrastructure, for now. They’re using it with customers, “but not in core banking operations yet due to high confidentiality and data protection rules,” Heiba says.

SOUND SMART- Non-bank lenders in Egypt like Klivvr and MNT are supervised by the Financial Regulatory Authority, which has in recent years been more willing to loosen rules and experiment with technology than the Central Bank of Egypt, which regulates the nation’s banks. The CBE didn’t reply to a request for comment for this story.

Egyptian banks aren’t spending much on AI, either. Only “a couple” of Egypt’s 30-plus banks are well ahead of the curve, BCG’s Fayek says. The nation’s two biggest state banks trail the region, too: on the Evident AI Index of 25 banks across the Middle East and Africa, Banque Misr ranked 19th and the National Bank of Egypt came last.

Computing power is another barrier: Central bank rules effectively prevent Egyptian lenders from sending customer data to cloud-based AI services like ChatGPT, Claude or Gemini, two bank CEOs confirmed to EnterpriseAM. Non-bank lenders have told us they feel similarly constrained. The result: Those experimenting with AI today are running smaller, open source models on their own servers instead, Abaza notes. That means buying GPUs, the super-expensive advanced computer chips that run most AI. “In Egypt, sometimes it’s very hard to get GPUs,” he says.

Anthropic files for blockbuster IPO

Anthropic filed on Monday for a Nasdaq IPO under the ticker ANT, setting up a test of investor appetite for AI as rivals OpenAI and Meta each seek tens of bns of USD of their own. Gulf investors that have already backed the maker of Claude will both get the chance to accumulate new shares and finally understand how public-market investors value their stock.

How much do Gulf shareholders really own? It’s anyone’s guess. The so-called S-1 filing runs more than 350 pages and some 240k words, but the only mention of MENA or Gulf countries relates to where securities can be marketed.

What about Gulf investment plans, spending, clients … anything? Nothing. Bupkis. Not a word.

That means that while big MENA+ institutions have significant stakes, none break the 5% threshold necessary to force disclosure. Known shareholders range from sovereign institutions (Abu Dhabi’s MGX, the Qatar Investment Authority, vehicles linked to Mubadala) to Damac, HOF Capital (founded by Onsi Sawiris, Hisham Elhaddad, and Fady Yacoub), Alpha Dhabi, and a unit of IHC. (See table, below.)

Notably absent: Saudi Arabia. Anthropic pooh-poohed bringing on Arab investors until recently and drew a line through suggestions that Saudi funds would be allowed into its cap table. Saudi’s Al Moammar Information Systems, a tiny shareholder, exited in 2025.

How much is the IPO worth? The filing leaves blank the price and number of shares on offer. Existing investors expect a USD 1.5-1.75 tn valuation, below the USD 2 tn-plus discussed in August, the FT reports, confirming an earlier account by Reuters. Trading is expected between the 3 November US midterms and Thanksgiving on 26 November. Look for the order book to open up next week.

Anthropic and its competitors are setting out to raise big money at the same time as investors show signs they may be cooling on AI as an asset class. The Nasdaq 100 has dipped a bit more than 7% from its 6 October high, edging nearer to the 10% drop needed to call it a market correction. Other would-be issuers are avoiding competition with Anthropic for institutional money, Reuters reports. There may also be some selling pressure from investors shedding existing holdings so they can write tickets for the Anthropic IPO, a pattern commonly seen before hot listings.

Other AI giants are feeling pressure

OpenAI is racing Anthropic. UAE investors and Coatue Management are discussing a round of up to USD 30 bn at a USD 1.25 tn pre-money valuation, CNBC reports. That target is down from an initial USD 1.4 tn. MGX, which co-led OpenAI’s USD 122 bn March round, is discussing another investment and additional UAE computing capacity could feature in a deal.

Meta, meanwhile, unveiled plans on Monday to raise up to USD 95 bn through a series of rolling stock offerings, also to fund growth of its AI offerings.

AND- Microsoft will offer buyouts to as much as 25% of its global workforce as it looks to keep pace with spiraling AI costs. As many as 60k people could be eligible to take a severance package, Reuters writes.

By the numbers

Anthropic’s sales pitch has substance: First-half revenue rose 1,300% y-o-y to USD 16.3 bn, while second-quarter operating profit came in at USD 974 mn. Its first-half operating loss narrowed to USD 693 mn from USD 2.1 bn a year earlier; full-year 2025 operating losses were USD 8.1 bn.

Its USD 80.7 bn first-half net loss isn’t as bad as it seems: It included USD 72.4 bn of non-cash charges from revaluing convertible notes and other instruments on the back of Anthropic’s rising valuation — it’s not a pure burn of USD 80.7 bn in cash from selling subsidized tokens. Net losses were USD 42 bn in 2025 and USD 1.5 bn in 2023.

IN CONTEXT- Anthropic’s half-year loss exceeds Fannie Mae’s USD 72 bn loss in 2009 and approaches AIG’s USD 99.3 bn in 2008 and AOL Time Warner’s USD 98.7 bn in 2002.

A whopping dose of concentration risk — and tons of cash on the balance sheet: Anthropic’s five biggest customers generated about 25% of revenue in the first nine months, although none individually exceeded 10%. The company was sitting on cash and cash-equivalents of USD 112.4 bn at the end of June 2026.

And Anthropic would have you believe it’s at least as worried about corporate inertia as it is about its competitors at OpenAI, Google, Microsoft, and Meta, writing in the risks section of the S-1 that corporate inertia is an obstacle to the quick adoption of Claude, which requires retraining staff, redesigning processes, and changing internal policies. Those obstacles are more pronounced in the regulated industries it most covets, from finance and law to healthcare.

AI is gonna kill us all? We never imagined we’d read a prospectus in which the issuer warns that its product might “pose catastrophic or existential risks to humanity” and concentrate power in ways that “erode trust, harm society, and destabilize the geopolitical order.”

AI is gonna save us? The S-1 also says that Anthropic’s aim is to “open new frontiers of discovery, and bring forth an age of miracles.”

ADVISORS- Goldman Sachs has the senior lead role alongside Morgan Stanley.

OpenAI seeks UAE backing as it looks to raise up to USD 30 bn

UAE investors are in talks to lead a fresh round of funding for OpenAI worth up to USD 30 bn alongside US investment firm Coatue, CNBC reports. The ChatGPT maker is said to be seeking a USD 1.25 tn valuation pre-money (financespeak for “before the new cash comes in”).

The competition for capital is going to be intense: Anthropic filed the S-1 for its blockbuster IPO yesterday and Meta is looking to raise c. USD 95 bn through a rolling series of stock offerings (see story, above).

OpenAI had initially pitched a valuation of c. USD 1.4 tn, but some investors balked, the Financial Times writes. The lower target would still comfortably blow past the company’s USD 852 bn valuation from March of this year.

BACKGROUND: MGX, backed by Mubadala and G42, co-led OpenAI’s USD 122 bn March round and is discussing another investment, as we noted last week. The firm also backs Anthropic.

Compute might also be on the table. OpenAI is looking for a deal on additional computing capacity in the UAE, CNBC says. Abu Dhabi officials would welcome it: The UAE’s broader play is to own pieces of major global AI labs, help build the infrastructure they need, and make certain that plenty of what they hope is future economic upside winds up captured at home.

IN CONTEXT- OpenAI seems to be getting its mojo back. September data from Ramp, a US provider of tech that allows businesses to manage spending on everything from AI to credit cards, show OpenAI adoption last month outpaced both Anthropic and the wider rate of AI adoption in general.

OpenAI has been on a tear of late. It has rolled out dots — AI agents that keep working between conversations — and Pages and Slides for collaborative work, with an Excel-like product in the works. Developers have gotten faster Codex options, expanded cloud coding, and security scanning, plus 28 days of new features or usage resets announced on X to win them over from Claude. The pitch, to tech nerds and knowledge workers alike: spend more of your working day inside OpenAI.

Middle East leads on data-center construction costs, report finds

The Middle East is the cheapest place in the world if you want to build a data center, according to Turner & Townsend’s 2026 construction cost index, released yesterday. It costs an average of USD 9.93 to build one watt of tech capacity here, against USD 12.06 in North America and USD 12.76 in Europe — or around 18% cheaper.

BY THE NUMBERS: The UAE comes in at USD 9.20, making it the fifth-cheapest of 62 markets surveyed. Qatar clocks in at USD 9.80 and Saudi Arabia at USD 10.80. Egypt was not included in the ranking.

Building an AI data center? You’ll pay more. Facilities specifically built for AI carry a construction premium of around 10% in the US, the report estimates, in large part due to the additional cost of cooling for more intensive loads.

The cost of building in the US is also driven up by the scale of the construction boom, which is making it difficult to lock in specialized contractors to build just about anything.

The Gulf starts from a cheaper base, but the report’s regional benchmarks cover conventional, air-cooled facilities, not the more expensive liquid-cooled systems needed in much of our part of the world. The comparisons also exclude land, utility works, and the computing equipment itself — including the chips.

It’s going to come down to availability (and cost) of power: 56% of industry respondents around the world identified power as the big obstacle to wrapping up a project. Nearly a third said it takes more than two years to get connected to utilities, Turner & Townsend says.

Gulf AI dealbook: Oman campuses, Alpha Dhabi funding, and Humain capacity

Oman has two new AI campuses on the drawing board. Hycap-backed UK green-energy developer Yamna has signed an MoU to advance a Salalah data center whose first phase is expected to cost more than GBP 1 bn, the UK government says. Yamna only started its data center business in April, but already has a separate green-ammonia project in Oman with two established players, EDF and Japan’s J-Power.

The bigger bet is a USD 3 bn campus that could be home to as much as 400 MW of compute down the road, led by Mubbadrah Investment, with Sustopia Capital, Vela Tech, and O2 Global Energy, TradeArabia reports. US AI infrastructure firm Crusoe — which counts Mubadala as an investor and which does a lot of business with OpenAI — has expressed interest in buying capacity, but the consortium hasn’t yet announced a signed offtake agreement.

MEANWHILE- Alpha Dhabi has agreed to commit USD 1 bn to Alpha Wave III, a global fund targeting AI and high-growth technology businesses, the Abu Dhabi investment group said yesterday. The commitment, which Alpha Dhabi said was still subject to unspecified “approvals,” will be funded with distributions from its earlier Alpha Wave investment and existing resources.

AND- Saudi Arabia’s Humain has a potential buyer for half of Oxagon’s first 100 MW worth of compute. Nasdaq-listed Nebius signed a preliminary agreement to take up to 50 MW from the Humain-DataVolt campus, with capacity expected to come online in phases starting 4Q 2027, our Saudi desk reported.

Who pays when AI agents use software?

Salesforce wants to bill your AI agent for using its software. The business software giant is targeting November to introduce agent registration and a new billing model, with an eye on charging you if you send agents into its software to pull reports or other data, according to a support notice. Salesforce hasn’t said how much it’s going to charge and isn’t yet tracking your usage.

The problem for software companies is simple: Customers might buy fewer seats for humans if they can send a bot to do more and more of the work. Charging for agent access could leave customers paying both their AI provider and the software company holding their data, The Information notes.

You can hear the software industry circling the wagons: “Headless is brainless,” Atlassian CEO Mike Cannon-Brookes told The Verge’s Nilay Patel on Decoder. He supports outside agents using Atlassian, but rejects giving up its interface altogether. That still leaves the buyer’s question: why pay for seats if the agent can do the work?

IN CONTEXT- Banks are also clamping down on Claude, Muse, and their robot buddies. JPMorgan Chase, Wells Fargo, Citi, Bank of America, and Morgan Stanley have all blocked or restricted outside agents’ access to customer accounts. Pundits speculate they fear the loss of low-cost deposits that constitute their feedstock.

Spyware AI, Cairo tech investment, and hospital agents lead regional roundup

The guy that helps intelligence services hack into iPhones now wants governments to trust his AI platform. Shalev Hulio, the co-founder of NSO, the Israeli maker of Pegasus spyware, is pitching country-hosted models through his new venture Dream as an alternative to US tech companies, the Financial Times reports. Dream’s investors include Abu Dhabi’s Tau Capital, backed by Sheikh Tahnoon bin Zayed Al Nahyan.

MEANWHILE- Deloitte is putting another USD 40 mn into its Cairo tech hub over the next 12–18 months, completing a USD 110 mn investment plan, as our Egypt desk reports. The unit exports AI, data, cloud, and cybersecurity services to clients in 30 countries.

M42, Mubadala and G42’s healthcare group, has deployed AI agents at seven UAE and Bahrain facilities. They draft notes, retrieve records, answer clinical questions, and suggest medical codes for staff review, the group says.

Abu Dhabi has also signed research deals with Isomorphic Labs, Whoop, and AstraZeneca covering AI drug discovery, wearables, and clinical trials, as our UAE desk reported.

Anthropic is widening its bid to capture consulting fees, reportedly agreeing to back a European consultancy with buyout firm Hg, the Financial Times writes, saying the firm could open offices in London and Berlin by year’s end. Engineers would work at client businesses to build AI products and help deploy the technology, putting Anthropic in direct competition with consultants including BCG and McKinsey as well as others selling its tools.

Apple’s next hardware reveal and the AI assistants you may not be able to use

Hold off on that MacBook purchase for a day. Apple has a big media event tomorrow, with an OLED touchscreen MacBook among the expected announcements, 9to5Mac reports. The device should also be lighter (but likely not much thinner) than the current model, Bloomberg’s Mark Gurman has speculated.

The iPhone Duo, meanwhile, went on sale over the weekend, starting at SAR 9,499 in Saudi — nearly SAR 10k before you’ve bought a case — and AED 8,499 in the UAE.

AND- You win some, you lose some: Apple unveiled a new iPad Mini yesterday. Win: It’s thinner, has an OLED screen, thinner bezels, and better sound. Loss: The screen maxes out at a 60 Hz refresh rate, not the 120 Hz “ProMotion” screen most of our fellow nerds were hoping for.

MEANWHILE- ChatGPT can now make a proper deck. Its new Slides feature lets Pro, Business, and Enterprise users with Space access edit and collaborate on presentations, then export them to PowerPoint, PDF, or Google Slides. It will not, sadly, make a 70-slide strategy presentation a good idea, and its controls are reminiscent of early Google Slides.

Watch this space: OpenAI is also teasing a Sheet application “coming soon” — at least in the “+ Create” menu of our Space tab on the Mac.app.

AND- OpenAI rolled out its GPT6.1 Terra model overnight, saying its cheapest model excels at visual reasoning and computer use. It also announced three new scientific discoveries (one each in mathematics, computer science, and physics) that are getting wide pickup on X.

That’s it for today! Like what you’re reading? Feel free to invite some friends. Or drop us a line at [email protected]. We’ll be back in your inbox on Thursday.

CALENDAR

October 2026

26-27 October (Monday-Tuesday): The Dubai AI Festival returns as part of Dubai AI Week.

28 October (Wednesday): Alphabet reports 3Q earnings after the US market closes, with AI capex and Gemini / cloud growth as the headline watch items.

28 October (Wednesday): Microsoft reports fiscal 1Q 2027 earnings after the US market closes, including Azure and OpenAI-partnership capex disclosures.

29 October (Thursday): Amazon reports 3Q earnings, including AWS and Bedrock AI growth.

30 October (Friday): UAE businesses with annual revenue of AED 50 mn or more must have an accredited e-invoicing provider in place by this date.

November 2026

7-10 November (Saturday-Tuesday): Cairo ICT runs in New Cairo alongside AIDC Expo, its dedicated AI, data center, and cloud event.

17-19 November (Tuesday-Thursday): Bengaluru Tech Summit brings businesses, startups, researchers, and policymakers to Bengaluru, India, under the theme “AI & Beyond.”

17-20 November (Tuesday-Friday): Microsoft Ignite runs in San Francisco and online, covering Microsoft’s AI, cloud, security, and data products.

30 November-4 December (Monday-Friday): AWS re:Invent brings Amazon’s cloud and AI conference to Las Vegas, with product announcements and sessions on deployment, costs, and security.

December 2026

1-3 December (Tuesday-Thursday): Black Hat MEA brings cybersecurity leaders, researchers, and technology companies to Malham, Riyadh.

6-12 December (Sunday-Saturday): NeurIPS holds its main AI research conference in Sydney, offering a look at advances that could shape the next generation of commercial AI.

7-11 December (Monday-Friday): GITEX Global opens with a leadership summit at Dubai World Trade Centre on 7 December, followed by its exhibition at Expo City Dubai on 8-11 December.

March 2027

23-24 March (Tuesday-Wednesday): Global AI Show Abu Dhabi

Sections

About

Privacy & Support