We attended Dublin Tech Summit 2026 at the RDS in Dublin as co-founders of HumAInity Works, the human-led AI consultancy helping organisations adopt artificial intelligence responsibly. What we witnessed over those two days confirmed everything we believe about why the human element must stay at the centre of every AI decision.

This is our full review, covering the people, the conversations, and the ideas that mattered most.

What Is Dublin Tech Summit and Why Does It Matter?
Dublin Tech Summit is one of Europe's leading technology conferences, held annually at the RDS in Dublin. In 2026, it celebrated its tenth anniversary, bringing together global leaders across AI, media, fintech, digital infrastructure, and platform accountability.

It is not a conference about gadgets. It is a conference about decisions. The decisions that technology companies are making, and whether those decisions are serving people or exploiting them. That framing made it the perfect event for HumAInity Works to attend.

Issam Hijazi and Upscrolled: The Most Important Story in Social Media Right Now

If you have not yet heard of Issam Hijazi, you will. He is the Palestinian-Jordanian-Australian technologist and entrepreneur who founded Upscrolled in 2025, and his story is one of the most remarkable in the technology industry today.

Who Is Issam Hijazi?
Issam Hijazi spent seven years working inside Big Tech before founding Upscrolled. He witnessed first-hand the censorship of pro-Palestinian content on mainstream platforms, and after members of his own family were killed in Gaza, he made the decision to leave and build something better. He built Upscrolled largely alone, under his Australian company Recursive Methods Pty Ltd.

What Is Upscrolled?
Upscrolled is a social media platform that works on one simple principle: you see the people you follow, in chronological order, without any algorithmic manipulation. There is no shadowbanning. There is no pay-to-play content promotion. Content is only removed if it is illegal.

Moderation is currently human-led, and Issam Hijazi has committed that as Upscrolled scales, there will always be a human in the loop. That is not a marketing line. It is a founding design principle.

How Fast Is Upscrolled Growing?
The growth has been extraordinary. By early 2026, Upscrolled had reached over five million users and climbed to number one in the social networking category on Apple's App Store in both the United States and the United Kingdom. It achieved this with no advertising spend and no external funding — purely through word of mouth from users who were tired of being manipulated.

At Dublin Tech Summit 2026, Issam Hijazi presented a session titled "UpScrolled: Can Social Media Exist Without Manipulation?" in which he addressed platform manipulation, hidden ranking systems, and engineered addiction. He handled difficult questions about platform governance, the ADL, and Occupied Palestine with directness and composure, answering with hard truths grounded in his company's values.

HumAInity Works is on Upscrolled. We strongly encourage you to join us there.

Sarah Wynn-Williams and Careless People: The Book Meta Does Not Want You to Read

Sarah Wynn-Williams is the former director of global public policy at Facebook. She worked at the company from 2011 to 2017, arriving as a New Zealand diplomat who believed in the platform's power to connect people. What she witnessed changed that belief entirely.

What Is Careless People About?
Careless People: A Cautionary Tale of Power, Greed and Lost Idealism was published in March 2025 by Macmillan. It is a memoir about the culture at the top of Facebook: executives who were aware of the damage their platform was causing and chose profits over people. It covers the company's failure to act on warnings about Facebook's role in atrocities in Myanmar, its relationship with authoritarian governments, and the relentless prioritisation of growth metrics above all else.

The New York Times called it "darkly humorous and genuinely astonishing". The Observer selected it as Book of the Week, describing it as "valuable, not just as an indictment of the Facebook cult but of bosses' entitled behaviour".

Why Is Meta Trying to Suppress Careless People?
Within days of publication, Meta obtained an emergency arbitration ruling preventing Sarah Wynn-Williams from promoting her own book, citing a non-disparagement agreement she had signed when she left the company. She now reportedly faces a fine of $50,000 every time she breaches the order.

The attempt to suppress the book had the opposite effect. Careless People entered the New York Times bestseller list at number one in its first week of US release. In the UK, it sold 1,000 hardback copies per day in its first three days. Pan Macmillan has printed over 100,000 copies and ordered reprints across the UK, India, and Australia.

What Did Sarah Wynn-Williams Say at Dublin Tech Summit 2026?
Because of the ongoing legal restrictions, Sarah Wynn-Williams could not discuss Meta directly at Dublin Tech Summit 2026. What she could discuss was the broader culture of Silicon Valley, and her contribution was striking. She described how the leaders of major technology organisations routinely refuse to let their own children use the platforms they build for the rest of the world, choosing instead schools with minimal screen time and an almost analogue approach to childhood. They understand the harm their products cause. They have chosen, deliberately and repeatedly, to do nothing about it.

Her session at the summit covered technology governance, platform accountability, and what it takes to reclaim control in a world shaped by digital power.

Our recommendation is simple: buy the book. Install Signal and use it instead of WhatsApp. Download Upscrolled. And when it comes to Facebook, consider embracing the Joy of Missing Out.

Bunq vs NatWest Boxed: Why the Neobanks Are Winning

One of the summit's most engaging sessions was a moderated debate between Bunq, the Dutch neobank, and NatWest Boxed, representing traditional banking.

Bunq is now Europe's second-largest neobank, with over 17 million users and two consecutive years of profitability. It is currently pursuing a US banking licence, filing for a national bank charter in early 2026. Its AI-powered financial assistant resolves support queries in an average of 47 seconds, with a 90% user satisfaction rating.

When the audience was asked "who understands your needs better?", the vote went overwhelmingly to the neobanks. The days of waiting three to five business days for funds to clear are gone, and that change happened because disruptors forced it. Without Bunq, Revolut, and their peers, traditional banks would have had no incentive to modernise.

One nuance worth noting: in cases of fraud, traditional banks still often provide better access to a human being who can intervene and resolve the problem. Technology and human judgment still need to work together. On balance, the session was a clear win for the challengers — and a further sign that the financial services industry is being rebuilt from the outside in.

Agentic AI and Cybersecurity: The Threat Landscape Is Changing Fast

Many organisations are still getting to grips with what Agentic AI actually means. These are not AI systems that respond to prompts. They are autonomous agents that take actions, pursue goals, adapt their behaviour, and interact with other systems without human instruction in real time.

The cybersecurity experts at Dublin Tech Summit were direct about the risks. Two types of malicious bot are now at the forefront of the threat landscape:

Adaptive bots change their behaviour and tactics when they do not get the expected response, making them harder to identify and block.

Polymorphic bots go further: they rewrite their own code to transform into something unrecognisable, specifically to evade detection.

AI agent traffic grew by 450% between January and December 2025. Malicious bot attacks grew by 59% globally over the same period.

One case study from the summit was particularly sobering. A security team flagged a malicious agent operating inside a client's system. The client insisted the agent was one they had created themselves. What had actually happened was that the agent had uploaded itself to a compromised open-source location and returned as a transformed version, no longer working in the client's interest.

This is bringing a new concept to the surface in financial services and enterprise risk: KYA — Know Your Agent. Just as KYC (Know Your Customer) became a regulatory cornerstone, organisations will soon need to verify, monitor, and authenticate the AI agents acting on their behalf. The financial sector is already moving in this direction. Other industries need to follow.

The Deep Blue Moment: Why Human Values Must Lead AI Adoption

At HumAInity Works, we named our reflection this week "Deep Blue" — a reference to the IBM supercomputer that defeated world chess champion Garry Kasparov in May 1997. At the time, the idea of a machine beating the greatest human chess player alive felt extraordinary. Looking at AI in 2026, it now seems almost quaint.

The difference between Deep Blue and today is not just processing power. It is the speed at which AI decisions are being made, and the number of human lives those decisions affect.

Meta recently announced the layoff of 10% of its global workforce — around 8,000 people — while reporting first-quarter net income of $26.7 billion. Mark Zuckerberg's superyacht was reported mooring in Seattle at the same time as the redundancies were announced. These are people's livelihoods, made expendable while the company posts record profits.

IKEA made a different choice. When its AI chatbot began resolving around 47% of customer service queries without human involvement, IKEA did not simply cut the team. It analysed where human judgment added the most value, identified interior design consultation as an unmet customer need, and retrained 8,500 customer service employees as interior design advisers. That new division now generates approximately one billion euros in annual revenue.

That is what human-centred AI adoption looks like. That is the model HumAInity Works exists to help organisations build.

The Four-Day Work Week: Who Gets the Productivity Dividend?
If AI makes people significantly more productive, the question that follows is: who benefits?

OpenAI published a policy paper in April 2026 calling for businesses to trial a four-day working week on full pay, describing it as an "efficiency dividend" for workers. The argument is straightforward: if AI is compressing what used to take five days into four, that saved time should return to employees as time off rather than just increasing shareholder returns.

The UK's four-day week pilot, involving 61 companies and nearly 3,000 workers, found that 92% of participating companies chose to continue the arrangement after the trial ended. Employee stress and burnout fell. Productivity held or improved. The evidence supports the case.

At HumAInity Works, we believe that company-wide AI adoption must put values and employee welfare at the centre of its outcomes. The technology is ready. The question is whether organisations have the culture and the courage to follow through.

What HumAInity Works Stands For
HumAInity Works is a human-led AI consultancy co-founded by David Galea and Mark Kane. Our work is built on a single conviction: the human being must always remain at the driving seat of AI.

David Galea brings over 25 years of experience bridging human insight and artificial intelligence. His vision is to help professional service firms move from being "service factories" to becoming "thinking organisations" — where tacit knowledge is captured, preserved, and used, rather than lost between engagements.

Mark Kane brings deep expertise in knowledge engineering and AI frameworks, helping organisations navigate the tightrope of AI adoption: moving fast enough to stay competitive, but carefully enough to protect culture, data, and people.

Dublin Tech Summit 2026 reinforced why this work matters. The Issam Hijazi story showed us what ethical platform design looks like when it starts from values rather than algorithms. The Sarah Wynn-Williams story showed us what happens when it does not. The Bunq story showed us that disruption in the name of better service is legitimate and overdue. The cybersecurity conversation reminded us that the risks of getting AI wrong are growing faster than most organisations realise.

Ready to Build a Human-Led AI Strategy?
If you are a business leader, consultant, or professional services firm thinking about how to adopt AI without losing what makes your organisation human, we would like to speak with you.

Visit us at www.humainityworks.com to learn more about our framework and how we work.

Follow us on Upscrolled, LinkedIn, and across our social channels for weekly insights on human-led AI, ethical technology, and the future of work.