Consumer protection Archives | Page 34 of 229

Structural friction, not intelligence, is holding back agentic AI

CIO leadership commentary highlights that many organisations investing in agentic AI, autonomous AI agents designed to execute complex, multi-step tasks, encounter disappointing results when deployments focus solely on outcomes like speed or cost savings without addressing underlying system design challenges.

The so-called ‘friction tax’ arises from siloed data, disjointed workflows and tools that force employees to act as manual connectors between systems, negating much of the theoretical efficiency AI promises.

The author proposes an ‘architecture of flow’ as a solution, in which context is unified across systems and AI agents operate on shared data and protocols, enabling work to move seamlessly between functions without bottlenecks.

This approach prioritises employee experience and customer value, enabling context-rich automation that reduces repetitive work and improves user satisfaction.

Key elements of such an architecture include universal context layers (e.g. standard protocols for data sharing) and agentic orchestration mechanisms that help specialised AI agents communicate and coordinate tasks across complex workflows.

When implemented effectively, this reduces cognitive load, strengthens adoption, and makes business growth a natural result of friction-free operations.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Enterprise AI security evolves as Cisco expands AI Defense capabilities

Cisco has announced a major update to its AI Defense platform as enterprise AI evolves from chat tools into autonomous agents. The company says AI security priorities are shifting from controlling outputs to protecting complex agent-driven systems.

The update strengthens end-to-end AI supply chain security by scanning third-party models, datasets, and tools used in development workflows. New inventory features help organisations track provenance and governance across AI resources.

Cisco has also expanded algorithmic red teaming through an upgraded AI Validation interface. The system enables adaptive multi-turn testing and aligns security assessments with NIST, MITRE, and OWASP frameworks.

Runtime protections now reflect the growing autonomy of AI agents. Cisco AI Defense inspects agent-to-tool interactions in real time, adding guardrails to prevent data leakage and malicious task execution.

Cisco says the update responds to the rapid operationalisation of AI across enterprises. The company argues that effective AI security now requires continuous visibility, automated testing, and real-time controls that scale with autonomy.

Would you like to learn more about AI, tech, and digital diplomacy? If so, ask our Diplo chatbot!

ChatGPT begins limited ads test in the US

OpenAI has begun testing advertisements inside ChatGPT for some adult users in the US, marking a major shift for the widely used AI service.

The ads appear only on Free and Go tiers in the US, while paid plans remain ad free. OpenAI says responses are unaffected, though critics warn commercial messaging could blur boundaries over time in the US.

Ads are selected based on conversation topics and prior interactions, prompting concern among privacy advocates in the US. OpenAI says advertisers receive only aggregated data and cannot view conversations.

Industry analysts say the move reflects growing pressure to monetise costly AI infrastructure in the US. Regulators and researchers continue to debate whether advertising can coexist with trust in AI systems.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

US lawsuits target social media platforms for deliberate child engagement designs

A landmark trial has begun in Los Angeles, accusing Meta and Google’s YouTube of deliberately addicting children to their platforms.

The case is part of a wider series of lawsuits across the US seeking to hold social media companies accountable for harms to young users. TikTok and Snap settled before trial, leaving Meta and YouTube to face the allegations in court.

The first bellwether case involves a 19-year-old identified as ‘KGM’, whose claims could shape thousands of similar lawsuits. Plaintiffs allege that design features were intentionally created to maximise engagement among children, borrowing techniques from slot machines and the tobacco industry.

A trial that may see testimony from executives, including Meta CEO Mark Zuckerberg, and could last six to eight weeks.

Social media companies deny the allegations, emphasising existing safeguards and arguing that teen mental health is influenced by numerous factors, such as academic pressure, socioeconomic challenges and substance use, instead of social media alone.

Meta and YouTube maintain that they prioritise user safety and privacy while providing tools for parental oversight.

Similar trials are unfolding across the country. New Mexico is investigating allegations of sexual exploitation facilitated by Meta platforms, while Oakland will hear cases representing school districts.

More than 40 state attorneys general have filed lawsuits against Meta, with TikTok facing claims in over a dozen states. Outcomes could profoundly impact platform design, regulation and legal accountability for youth-focused digital services.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

AI governance takes focus at UN security dialogue

The UN will mark the fourth International Day for the Prevention of Violent Extremism Conducive to Terrorism on 12 February 2026 with a high-level dialogue focused on AI. The event will examine how emerging technologies are reshaping both prevention strategies and extremist threats.

Organised by the UN Office of Counter-Terrorism in partnership with the Republic of Korea’s UN mission, the dialogue will take place at UN Headquarters in New York. Discussions will bring together policymakers, technology experts, civil society representatives, and youth stakeholders.

A central milestone will be the launch of the first UN Practice Guide on Artificial Intelligence and Preventing and Countering Violent Extremism. The guide offers human rights-based advice on responsible AI use, addressing ethical, governance, and operational risks.

Officials warn that AI-generated content, deepfakes, and algorithmic amplification are accelerating extremist narratives online. Responsibly governed AI tools could enhance early detection, research, and community prevention efforts.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

EU challenges Meta over WhatsApp AI restrictions

The European Commission has warned Meta that it may have breached EU antitrust rules by restricting third-party AI assistants from operating on WhatsApp. A Statement of Objections outlines regulators’ preliminary view that the policy could distort competition in the AI assistant market.

The probe centres on updated WhatsApp Business terms announced in October 2025 and enforced from January 2026. Under the changes, rival general-purpose AI assistants were effectively barred from accessing the platform, leaving Meta AI as the only integrated assistant available to users.

Regulators argue that WhatsApp serves as a critical gateway for consumers AI access AI services. Excluding competitors could reinforce Meta’s dominance in communication applications while limiting market entry and expansion opportunities for smaller AI developers.

Interim measures are now under consideration to prevent what authorities describe as potentially serious and irreversible competitive harm. Meta can respond before any interim measures are imposed, while the broader antitrust probe continues.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

EU faces pressure to boost action on health disinformation

A global health organisation is urging the EU to make fuller use of its digital rules to curb health disinformation as concerns grow over the impact of deepfakes on public confidence.

Warnings point to a rising risk that manipulated content could reduce vaccine uptake instead of supporting informed public debate.

Experts argue that the Digital Services Act already provides the framework needed to limit harmful misinformation, yet enforcement remains uneven. Stronger oversight could improve platforms’ ability to detect manipulated content and remove inaccurate claims that jeopardise public health.

Campaigners emphasise that deepfake technology is now accessible enough to spread false narratives rapidly. The trend threatens vaccination campaigns at a time when several member states are attempting to address declining trust in health authorities.

The EU officials continue to examine how digital regulation can reinforce public health strategies. The call for stricter enforcement highlights the pressure on Brussels to ensure that digital platforms act responsibly rather than allowing misleading material to circulate unchecked.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Discord expands teen-by-default protection worldwide

Discord is preparing a global transition to teen-appropriate settings that will apply to all users unless they confirm they are adults.

The phased rollout begins in early March and forms part of the company’s wider effort to offer protection tailored to younger audiences rather than relying on voluntary safety choices. Controls will cover communication settings, sensitive content and access to age-restricted communities.

The update is based on an expanded age assurance system designed to protect privacy while accurately identifying users’ age groups. People can use facial age estimation on their own device or select identity verification handled by approved partners.

Discord will also rely on an age-inference model that runs quietly in the background. Verification results remain private, and documents are deleted quickly, with users able to appeal group assignments through account settings.

Stricter defaults will apply across the platform. Sensitive media will stay blurred unless a user is confirmed as an adult, and access to age-gated servers or commands will require verification.

Message requests from unfamiliar contacts will be separated, friend-request alerts will be more prominent and only adults will be allowed to speak on community stages instead of sharing the feature with teens.

Discord is complementing the update by creating a Teen Council to offer advice on future safety tools and policies. The council will include up to a dozen young users and aims to embed real teen insight in product development.

The global rollout builds on earlier launches in the UK and Australia, adding to an existing safety ecosystem that includes Teen Safety Assist, Family Centre, and several moderation tools intended to support positive and secure online interactions.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Study questions reliability of AI medical guidance

AI chatbots are not yet capable of providing reliable health advice, according to new research published in the journal Nature Medicine. Findings show users gain no greater diagnostic accuracy from chatbots than from traditional internet searches.

Researchers tested nearly 1,300 UK participants using ten medical scenarios, ranging from minor symptoms to conditions requiring urgent care. Participants were assigned to use either OpenAI’s GPT-4o, Meta’s Llama 3, Command R+, or a standard search engine to assess symptoms and determine next steps.

Chatbot users identified their condition about one-third of the time, with only 45 percent selecting the correct medical response. Performance levels matched those relying solely on search engines, despite AI systems scoring highly on medical licensing benchmarks.

Experts attributed the gap to communication failures. Users often provided incomplete information or misinterpreted chatbot guidance.

Researchers and bioethicists warned that growing reliance on AI for medical queries could pose public health risks without professional oversight.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!

Crypto exchange scrambles after $40bn Bitcoin payout error

South Korea’s second-largest cryptocurrency exchange, Bithumb, is attempting to recover more than $40bn in Bitcoin after a promotional payout error credited customers with Bitcoin rather than Korean won.

The mistake occurred on 6 February during a ‘random box’ event, when prize values were entered in Bitcoin rather than in Bitcoin. Intended rewards totalled 620,000 won for 695 users, yet 620,000 bitcoins were distributed.

Only 249 customers opened their boxes, but the credited sums exceeded the exchange’s holdings.

Most balances were reversed through internal ledger corrections. About 13bn won ($9m) remains unrecovered after some users sold or withdrew funds before accounts were frozen. Authorities said 86 customers liquidated roughly 1,788 Bitcoins within 35 minutes.

Regulators have opened a full investigation, and lawmakers have scheduled an emergency hearing. Legal uncertainty remains over liability, while the exchange confirmed no hacking was involved and pledged stronger internal controls.

Would you like to learn more about AI, tech and digital diplomacy? If so, ask our Diplo chatbot!