Artificial intelligence emerged as a bigger concern than nuclear weapons during a major strategic stability discussion at the Shangri-La Dialogue in Singapore, as senior military leaders warned that AI is accelerating the speed of conflict beyond the pace of human decision-making.

Why AI Dominated The Discussion

The Shangri-La Dialogue is one of the world’s most important defence and security forums, bringing together defence ministers, military commanders, policymakers, and security experts from across the Indo-Pacific and beyond.

Traditionally, discussions about strategic stability have focused heavily on nuclear deterrence, missile defence, arms control, and the balance of power between major states. This year, however, much of the conversation centred on artificial intelligence and the risks it could introduce into military decision-making.

The concern was not that AI is becoming conscious or uncontrollable. Rather, military leaders repeatedly highlighted the possibility that AI-enabled systems could compress decision-making timelines so dramatically that humans struggle to assess situations properly before responding.

Lieutenant General Nauman Zakria, Commander of 1 Corps and the Army Rocket Force Command of the Pakistan Army, explained the problem through the military concept known as the OODA loop, which stands for Observe, Orient, Decide, and Act.

According to Zakria, AI is compressing that cycle to the point where “a human can’t evaluate the situation fast enough.”

He warned that under such conditions, “People will act irrationally, and the actions will be extreme.”

Why Military Leaders Are So Worried

The concern centres on escalation. For example, for decades, strategic stability has relied on the assumption that leaders have sufficient time to evaluate information, consult advisers, communicate with allies, and consider the consequences of military action.

AI-enabled systems have the potential to analyse data, identify threats, recommend responses, and support operational decisions far faster than human decision-makers can process information themselves.

That speed may offer significant military advantages. However, crucially, it also raises the possibility that misunderstandings, false alarms, technical errors, or incorrect threat assessments could trigger responses before humans have time to intervene.

The faster the decision cycle becomes, the less opportunity there is to question assumptions or prevent mistakes.

Already On The Battlefield

Several speakers stressed that these concerns are no longer theoretical. For example, General Onno Eichelsheim, Chief of Defence of the Netherlands, pointed to the growing use of AI in active conflicts, including Ukrainian efforts to anticipate Russian attacks and coordinate drone operations. The United States has also confirmed using AI tools to support military planning and targeting decisions.

As Eichelsheim put it: “AI is a huge risk in escalation. I think that’s clear.”

At the same time, he acknowledged that military adoption is unlikely to slow, adding: “But I’m not naive. It’ll be used in the domain. It is already being used.”

Those comments reflect a growing reality that military organisations increasingly see AI as a capability they cannot afford to ignore, even while recognising the risks associated with it.

The Pace Is Accelerating

What makes these concerns more pressing is the speed at which military AI capabilities are evolving. Increasingly, AI is being integrated with drones, autonomous systems, and robotics rather than operating solely as a decision-support tool.

For example, US startup Foundation Future Industries, which has Eric Trump as an adviser, recently tested humanoid robots in Ukraine with support from Ukrainian authorities. The company says it hopes to deploy more advanced versions with military forces within the next 12 to 18 months. Although the trials focused on logistics rather than combat, they illustrate how AI, robotics, and military operations are becoming increasingly interconnected.

Taken together, these developments suggest that AI is moving beyond the planning room and becoming a more active part of military operations, raising further questions about human oversight, accountability, and control during conflict.

A Humanitarian Concern

The strongest warning came from Mirjana Spoljaric, President of the International Committee of the Red Cross. While military leaders focused largely on strategic and operational risks, Spoljaric highlighted the humanitarian implications of increasingly automated warfare.

She warned that the growing distance between decision-makers and the battlefield creates new challenges for accountability and civilian protection.

“We don’t know where the trigger is pulled,” she said. “It could be thousands of kilometres away.”

Spoljaric also argued that while AI may eventually offer benefits for civilian protection, current deployments are highlighting more of the risks than the advantages.

Her comments reflect wider concerns among humanitarian organisations that meaningful human control over lethal decisions could become increasingly difficult to maintain as AI capabilities advance.

Nuclear Weapons Have Not Gone Away

It should be noted here, however, that nuclear deterrence remained part of the discussion. Major General Meng Xiangqing of China’s People’s Liberation Army reaffirmed China’s long-standing no-first-use nuclear policy and proposed broader commitments among nuclear powers.

“If we can do so, we can reduce the risk and we can further enhance strategic stability,” he said. However, the fact that AI repeatedly returned to the centre of the discussion was itself notable.

A panel that would traditionally be dominated by nuclear strategy spent much of its time examining how artificial intelligence could reshape military escalation, crisis management, and conflict decision-making.

What Does This Mean For Your Business?

For businesses, the discussion highlights how quickly AI is moving from a productivity and automation tool into an issue of national security, geopolitics, and strategic risk.

Many organisations currently view AI primarily through the lens of efficiency, customer service, software development, or data analysis. Governments and defence planners are increasingly focused on a different question, namely what happens when AI systems begin influencing decisions where mistakes carry serious consequences.

The wider significance is that AI’s impact may ultimately extend far beyond individual applications or business processes. The technology is beginning to affect how countries think about defence, deterrence, crisis management, and international stability.

The message emerging from the Shangri-La Dialogue was not that AI should be stopped. Rather, it was that societies may need to think much more carefully about how much decision-making authority is delegated to systems that can operate faster than humans can fully understand the situations they are responding to.

Microsoft has drawn criticism from parts of the cyber security community after publicly condemning a researcher who disclosed several unpatched Windows vulnerabilities and warning that its Digital Crimes Unit would continue pursuing those who enable criminal activity.

What Happened?

The dispute centres on a researcher known online as “Nightmare Eclipse”, who recently published proof-of-concept exploit code for a series of vulnerabilities affecting Microsoft Defender and BitLocker.

The flaws, which became known as BlueHammer, RedSun, UnDefend, YellowKey, GreenPlasma, and MiniPlasma, were disclosed publicly before Microsoft had released patches for all of them. Some have since been assigned CVE identifiers, while Microsoft and the US Cybersecurity and Infrastructure Security Agency (CISA) have confirmed that certain vulnerabilities have been exploited in real-world attacks.

Microsoft argues that the disclosures put customers at unnecessary risk because the company was not given sufficient opportunity to investigate and fix the flaws before exploit code became publicly available.

In a post published by the Microsoft Security Response Center, the company said: “The details of these vulnerabilities were not shared with Microsoft prior to release, and the disclosures put our customers at unnecessary risk.”

The company also stated: “Uncoordinated disclosures that put proof-of-concept code for unpatched vulnerabilities into the hands of bad actors are never justifiable and have real-world consequences.”

Why The Response Has Proved Controversial

The strongest reaction did not come from Microsoft’s criticism of the disclosures themselves, but from language used elsewhere in the company’s statement.

Microsoft wrote that its Digital Crimes Unit “will continue bringing cases against these actors and those that enable their criminal activity – coordinating as needed with law enforcement around the world.”

Many researchers interpreted that wording as a threat directed at vulnerability researchers, particularly given the public nature of the dispute.

The controversy intensified because Nightmare Eclipse claims to have previously attempted to engage with Microsoft through its Microsoft Security Response Center process before later having their account revoked. Microsoft has not publicly addressed those specific claims.

The researcher subsequently published the vulnerabilities through GitHub and GitLab, and their accounts on both platforms have since been removed.

A Debate That Has Been Running For Decades

The dispute touches on one of cyber security’s longest-running arguments about how vulnerabilities should be disclosed.

For example, most of the industry now follows what is known as Coordinated Vulnerability Disclosure, or CVD. Under this model, researchers privately notify software vendors about security flaws, give them time to investigate and develop fixes, and then publish technical details once patches become available.

Microsoft strongly supports that approach. In its statement, the company described CVD as “the industry standard” and said it works with “hundreds of security researchers” each year through the process.

However, critics argue that disclosure relationships only work when vendors respond quickly, communicate effectively, and treat researchers fairly. When researchers believe their concerns are being ignored, disputes can arise over whether public disclosure becomes justified.

The disagreement is significant because independent researchers play a major role in identifying vulnerabilities that software vendors might otherwise miss.

Why This Matters Beyond Microsoft

The row has happened at a time when vulnerability discovery is accelerating across the industry. For example, recent advances in AI-assisted security research are enabling researchers and organisations to identify flaws at unprecedented speed. At the same time, software suppliers are facing growing backlogs of vulnerabilities to investigate, validate, and patch.

That creates tension on both sides. Vendors want time to protect customers before details become public. Researchers want assurance that their findings will be taken seriously and addressed promptly.

The result is growing pressure on disclosure processes that were designed for a slower era of software development and vulnerability discovery.

The wider concern expressed by many researchers is that aggressive responses to disclosure disputes could discourage future reporting.

Microsoft itself acknowledged the importance of the research community, stating: “Our team will continue to support responsible research as we do everything we can to quickly investigate, address, and release updates for vulnerabilities that impact our customers.”

The company also said: “We always have and will continue to welcome vulnerability submissions from anyone through our public researcher portal, regardless of past interactions or reputation.”

What Does This Mean For Your Business?

For businesses, the most important issue here is not really the disagreement itself but the vulnerabilities at the centre of it.

The Defender and BitLocker flaws highlight how even widely trusted security tools can contain weaknesses that attackers may seek to exploit. Organisations should therefore continue prioritising patch management, endpoint monitoring, vulnerability scanning, and defence-in-depth controls rather than assuming any single security product provides complete protection.

The wider lesson is that the relationship between software vendors and independent researchers remains an essential part of cyber security. Vulnerabilities are often discovered by external researchers long before vendors become aware of them, making cooperation between the two groups critical to keeping systems secure.

Whether Microsoft handled this particular dispute correctly will continue to be debated. However, most security professionals would at least agree that a disclosure process that encourages researchers to report vulnerabilities and vendors to fix them quickly remains one of the most important defences the industry has.

Literary agents are increasingly being impersonated by scammers as AI makes it easier to create convincing fake identities, websites, emails, and publishing credentials.

Mark Gottlieb of Trident Media Group says artificial intelligence has not simply accelerated publishing fraud but has “industrialised it”. Scammers can now quickly create realistic agency websites, correspondence, and professional profiles.

The scams exploit the fact that many aspiring authors have limited experience working with literary agents and often communicate remotely. Fraudsters can therefore pose as legitimate agents and offer representation or publishing opportunities.

The threat is also evolving beyond upfront fees. Stolen manuscripts can be turned into counterfeit ebooks, fake audiobooks, unauthorised translations, or AI-generated derivative works.

Businesses and individuals should verify identities through official channels, check websites and email domains carefully, and be cautious of unexpected approaches. As AI makes impersonation more convincing, confirming who you are dealing with is becoming an increasingly important security measure.

The European Commission is encouraging households to move electricity consumption away from peak periods as rising demand from AI data centres, electrification, and digital infrastructure places growing pressure on Europe’s power grids.

What The EU Has Announced

As part of its new Strategic Roadmap for Digitalisation and Artificial Intelligence in Energy, the European Commission has outlined plans to accelerate the rollout of smart meters and other digital technologies designed to help consumers use electricity when demand is lower and prices are cheaper.

The initiative forms part of a broader effort to modernise Europe’s energy system while managing rapidly growing electricity demand.

Alongside the roadmap, the Commission has also introduced a Data Centre Energy Efficiency Package that includes a new rating scheme for data centres and lays the groundwork for future minimum energy performance standards.

According to the Commission, digital solutions can help consumers “shift consumption to hours when electricity is cheaper and thereby lower their energy bills.”

The Commission believes that greater demand-side flexibility could reduce electricity costs for EU consumers by more than €71 billion per year.

Why Data Centres Are Becoming Part Of The Energy Debate

The growing focus on electricity demand is closely linked to the rapid expansion of AI infrastructure.

Training and operating advanced AI models requires vast computing resources, much of which is housed in large-scale data centres. As AI adoption accelerates, so does the amount of electricity needed to power and cool those facilities.

According to the Commission, data centres already account for around 2.5 per cent of EU electricity consumption, and demand is expected to more than double over the next four years.

At the same time, electricity demand is also increasing from electric vehicles, heat pumps, hydrogen production, and the wider electrification of the economy.

The result is a growing challenge for policymakers attempting to balance economic growth, climate goals, energy security, and affordability.

Ireland Offers A Glimpse Of The Challenge

Ireland provides one of the clearest examples of the pressures that can emerge when data centre growth outpaces energy infrastructure investment.

Data centres now consume more than 22 per cent of Ireland’s national electricity supply, making it one of the most concentrated data centre markets in the world.

The issue has become significant enough that some proposed developments have faced planning and grid-capacity challenges. Concerns have also been raised about the potential impact on electricity prices in regions with large concentrations of digital infrastructure.

While AI data centres are not the sole cause of rising energy demand, they are becoming an increasingly visible contributor to a broader capacity challenge affecting many countries.

A Difficult Balancing Act

The situation highlights a growing tension within European policy. For example, on one hand, the EU wants to accelerate AI development and reduce dependence on foreign technology providers. On the other, the infrastructure required to support those ambitions consumes large amounts of electricity at a time when Europe is simultaneously trying to decarbonise its economy and keep energy affordable.

The Commission argues that digitalisation can help address part of the problem. The roadmap notes that AI-based optimisation of energy systems could improve efficiency, reduce waste, and make better use of existing infrastructure.

As the Commission states, “Tech sovereignty in the energy sector is therefore more urgent than ever” while digital technologies can help create “a clean, competitive and secure EU energy system.”

However, efficiency improvements alone may not solve the underlying challenge if electricity demand continues to grow faster than generation and grid capacity.

What Does This Mean For Your Organisation?

For organisations, the announcement highlights a sustainability issue that is likely to become increasingly important over the next decade.

AI offers significant opportunities for innovation, productivity, and economic growth. However, the infrastructure required to support those benefits has real environmental and energy consequences that governments, businesses, and consumers will need to manage.

The Commission’s response suggests that future energy policy may focus not only on generating more electricity but also on using existing capacity more intelligently through smart meters, AI-enabled grid management, demand flexibility, and stricter efficiency standards.

The wider lesson is that the sustainability debate around AI is moving beyond questions about individual technologies and towards a much larger discussion about how societies generate, distribute, and consume energy in an increasingly digital world.

The new Copilot PowerPoint agent can turn a simple prompt into a fully structured presentation, helping you create polished slides in minutes while reducing manual effort and saving time.

[Note – To Watch This Video without glitches/interruptions, It may be best to download it first]

Google Lens in Chrome can extract text directly from paused videos, webinars, Teams recordings, training videos, and online presentations, making it easy to copy URLs, settings, serial numbers, product codes, or notes without having to type them manually.

Why It Works

Most people associate Google Lens with photographs, but it can also analyse any image displayed in your browser, including individual frames from videos.

This means you can quickly capture information displayed on screen, even if the video creator hasn’t provided it elsewhere.

It’s particularly useful for copying technical settings, website addresses, software licence information, presentation content, contact details, or troubleshooting instructions from training videos and webinars.

How To Use It

1. Open the video in Google Chrome.

2. Pause the video on the frame containing the text you want to copy.

3. Right-click the video frame.

4. Select ‘Search Image With Google Lens’.

5. When the Lens panel opens, select the ‘Text’ option.

6. Highlight the text and click ‘Copy Text’.

A Few Things To Remember

Google Lens works best when the text is clear and reasonably large on screen.

It can be used with YouTube videos, online training platforms, recorded Teams meetings, webinars, presentations, screenshots, and many other types of visual content.

For business users, it’s a simple way to save time and avoid mistakes when copying information from video-based training or support materials.