Content Policy

DeepMind unveils AlphaEvolve for scientific breakthroughs

Google DeepMind has unveiled AlphaEvolve, a new AI system designed to help solve complex scientific and mathematical problems by improving how algorithms are developed.

Rather than acting like a standard chatbot, AlphaEvolve blends large language models from the Gemini family with an evolutionary approach, enabling it to generate, assess, and refine multiple solutions at once.

Instead of relying on a single output, AlphaEvolve allows researchers to submit a problem and potential directions. The system then uses both Gemini Flash and Gemini Pro to create various solutions, which are automatically evaluated.

The best results are selected and enhanced through an iterative process, improving accuracy and reducing hallucinations—a common issue with AI-generated content.

Unlike earlier DeepMind tools such as AlphaFold, which focused on narrow domains, AlphaEvolve is a general-purpose AI for coding and algorithmic tasks.

It has already shown its value by optimising Google’s own Borg data centre management system, delivering a 0.7% efficiency gain—significant given Google’s global scale.

The AI also devised a new method for multiplying complex matrices, outperforming a decades-old technique and even beating DeepMind’s specialised AlphaTensor model.

AlphaEvolve has also contributed to improvements in Google’s hardware design by optimising Verilog code for upcoming Tensor chips.

Though not publicly available yet due to its complexity, AlphaEvolve’s evaluation-based framework could eventually be adapted for smaller AI tools used by researchers elsewhere.

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

Kick faces investigation after ignoring Ofcom risk assessment request

Ofcom has launched two investigations into Kick Online Entertainment, the provider of a pornography website, over potential breaches of the Online Safety Act.

The regulator said the company failed to respond to a statutory request for a risk assessment related to illegal content appearing on the platform.

As a result, Ofcom is investigating whether Kick has failed to meet its legal obligations to complete and retain a record of such a risk assessment, as well as for not responding to the regulator’s information request.

Ofcom confirmed it had received complaints about potentially illegal material on the site, including child sexual abuse content and extreme pornography.

It is also considering a third investigation into whether the platform has implemented adequate safety measures to protect users from such material—another requirement under the Act.

Under the Online Safety Act, firms found in breach can face fines of up to £18 million or 10% of their global revenue, whichever is higher. In the most severe cases, Ofcom can pursue court orders to block UK access to the website or compel payment providers and advertisers to cut ties with the platform.

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

Meta targets critics as FTC case unfolds

Long-standing friction between Big Tech and the media resurfaced during Meta’s antitrust trial with the Federal Trade Commission this week. In a heated courtroom exchange, Meta’s legal team used critical commentary from prominent tech journalists to cast doubt on the FTC’s case.

Meta’s lead attorney, Mark Hansen, questioned the credibility of FTC expert Scott Hemphill by referencing a 2019 antitrust pitch Hemphill co-authored with Facebook co-founder Chris Hughes and former White House advisor Tim Wu.

The presentation cited public statements from reporters Kara Swisher and Om Malik as evidence of Meta’s dominance and aggressive acquisitions.

Hansen dismissed Malik as a ‘failed blogger’ with personal bias and accused Swisher of similar hostility, projecting a headline where she described Mark Zuckerberg as a ‘small little creature with a shriveled soul.’

He also attempted to discredit a cited New York Post article by invoking the tabloid’s notorious ‘Headless Body in Topless Bar’ cover.

These moments highlight Meta’s growing resentment toward the press, which has intensified alongside rising criticism of its business practices. Once seen as scrappy disruptors, Facebook and other tech giants now face regular scrutiny—and appear eager to push back.

Swisher and Malik have both openly criticized Meta in the past. Swisher famously challenged Zuckerberg over content moderation and political speech, while Malik has questioned the company’s global expansion strategies.

Their inclusion in a legal document presented in court underscores how media commentary is influencing regulatory narratives. Meta has previously blamed critical press for damaging user sentiment in the wake of scandals like Cambridge Analytica.

The FTC argues that consistent engagement levels despite bad press prove Meta’s monopoly power—users feel they have no real alternatives to Facebook and Instagram. As the trial continues, so too does Meta’s public battle—not just with regulators, but with the journalists documenting its rise and reckoning.

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

Instagram calls for EU-wide teen protection rules

Instagram is calling on the European Union to introduce new regulations requiring app stores to implement age verification and parental approval systems.

The platform argues that such protections, applied consistently across all apps, are essential to safeguarding teenagers from harmful content online.

‘The EU needs consistent standards for all apps, to help keep teens safe, empower parents and preserve privacy,’ Instagram said in a blog post.

The company believes the most effective way to achieve this is by introducing protections at the source—before teenagers download apps from the Apple App Store or Google Play Store.

Instagram is proposing that app stores verify users’ ages and require parental approval for teen app downloads. The social media platform cites new research from Morning Consult showing that three in four parents support such legislation.

Most parents also view app stores, rather than individual apps, as the safer and more manageable point for controlling what their teens can access.

To reinforce its position, Instagram points to its own safety efforts, such as the introduction of Teen Accounts. These private-by-default profiles limit teen exposure to messages and content from unknown users, and apply stricter filters to reduce exposure to sensitive material.

Instagram says it is working with civil society groups, industry partners, and European policymakers to push for rules that protect young users across platforms. With teen safety a growing concern, the company insists that industry-wide, enforceable solutions are urgently needed.

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

Republicans seek to block state AI laws for a decade

Republican lawmakers in the US have introduced a proposal that would block states from regulating artificial intelligence for the next ten years. Critics argue the move is a handout to Big Tech and could stall protections already passed in states like California, Utah, and Colorado.

The measure, embedded in a budget reconciliation bill, would prevent states from enforcing rules on a wide range of automated systems, from AI chatbots to algorithms used in health and justice sectors.

Over 500 AI-related bills have been proposed this year at the state level, and many of them would be nullified if the federal ban succeeds. Supporters of the bill claim AI oversight should happen at the national level to avoid a confusing patchwork of state laws.

Opponents, including US Democrats and tech accountability groups, warn the ban could allow unchecked algorithmic discrimination, weaken privacy, and leave the public vulnerable to AI-driven harms.

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

TikTok unveils AI video feature

TikTok has launched ‘AI Alive,’ its first image-to-video feature that allows users to transform static photos into animated short videos within TikTok Stories.

Accessible only through the Story Camera, the tool applies AI-driven movement and effects—like shifting skies, drifting clouds, or expressive animations—to bring photos to life.

Unlike text-to-image tools found on Instagram and Snapchat, TikTok’s latest feature takes visual storytelling further by enabling full video generation from single images. Although Snapchat plans to introduce a similar function, TikTok has moved ahead with this innovation.

All AI Alive videos will carry an AI-generated label and include C2PA metadata to ensure transparency, even when shared beyond the platform.

TikTok emphasises safety, noting that every AI Alive video undergoes several moderation checks before it appears to creators.

Uploaded photos, prompts, and generated videos are reviewed to prevent rule-breaking content. Users can report violations, and final safety reviews are conducted before public sharing.

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

Harvey adds Google and Anthropic AI

Harvey, the fast-growing legal AI startup backed early by the OpenAI Startup Fund, is now embracing foundation models from Google and Anthropic instead of relying solely on OpenAI’s.

In a recent blog post, the company said it would expand its AI model options after internal benchmarks showed that different tools excel at different legal tasks.

The shift marks a notable win for OpenAI’s competitors, even though Harvey insists it’s not abandoning OpenAI. Its in-house benchmark, BigLaw, revealed that several non-OpenAI models now outperform Harvey’s original system on specific legal functions.

For instance, Google’s Gemini 2.5 Pro performs well at legal drafting, while OpenAI’s o3 and Anthropic’s Claude 3.7 Sonnet are better suited for complex pre-trial work.

Instead of building its own models, Harvey now aims to fine-tune top-tier offerings from multiple vendors, including through Amazon’s cloud. The company also plans to launch a public legal benchmark leaderboard, combining expert legal reviews with technical metrics.

While OpenAI remains a close partner and investor, Harvey’s broader strategy signals growing competition in the race to serve the legal industry with AI.

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

Masked cybercrime groups rise as attacks escalate worldwide

Cybercrime is thriving like never before, with hackers launching attacks ranging from absurd ransomware demands of $1 trillion to large-scale theft of personal data. Despite efforts from Microsoft, Google and even the FBI, these threat actors continue to outpace defences.

A new report by Group-IB has analysed over 1,500 cybercrime investigations to uncover the most active and dangerous hacker groups operating today.

Rather than fading away after arrests or infighting, many cybercriminal gangs are re-emerging stronger than before.

Group-IB’s May 2025 report highlights a troubling increase in key attack types across 2024 — phishing rose by 22%, ransomware leak sites by 10%, and APT (advanced persistent threat) attacks by 58%. The United States was the most affected country by ransomware activity.

At the top of the cybercriminal hierarchy now sits RansomHub, a ransomware-as-a-service group that emerged from the collapsed ALPHV group and has already overtaken long-established players in attack numbers.

Behind it is GoldFactory, which developed the first iOS banking trojan and exploited facial recognition data. Lazarus, a well-known North Korean state-linked group, also remains highly active under multiple aliases.

Meanwhile, politically driven hacktivist group NoName057(16) has been targeting European institutions using denial-of-service attacks.

With jurisdictional gaps allowing cybercriminals to flourish, these masked hackers remain a growing concern for global cybersecurity, especially as new threat actors emerge from the shadows instead of disappearing for good.

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

US Copyright Office avoids clear decision on AI and fair use

The US Copyright Office has stopped short of deciding whether AI companies can legally use copyrighted material to train their systems under fair use.

Its newly released report acknowledges that some uses—such as non-commercial research—may qualify, while others, like replicating expressive works from pirated content to produce market-ready AI output, likely won’t.

Rather than offering a definitive answer, the Office said such cases must be assessed by the courts, not through a universal standard.

The latest report is the third in a series aimed at guiding how copyright law applies to AI-generated content. It reiterates that works entirely created by AI cannot be copyrighted, but human-edited outputs might still qualify.

The 108-page document focuses heavily on whether AI training methods transform content enough to justify legal protection, and whether they harm creators’ livelihoods through lost sales or diluted markets.

Instead of setting new policy, the Office highlights existing legal principles, especially the four factors of fair use: the purpose, the nature of the work, the amount used, and the impact on the original market.

It notes that AI-generated content can sometimes alter original works meaningfully, but when styles or outputs closely resemble protected material, legal risks remain. Tools like content filters are seen as helpful in preventing infringement, even though they’re not always reliable.

The timing of the report has been overshadowed by political turmoil. President Donald Trump reportedly dismissed both the Librarian of Congress and the head of the Copyright Office days before the report’s release.

Meanwhile, creators continue urging the government not to permit fair use in AI training, arguing it threatens the value of original work. The debate is now expected to unfold further in courtrooms instead of regulatory offices.

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

Jamie Lee Curtis calls out Zuckerberg over AI scam using her likeness

Jamie Lee Curtis has directly appealed to Mark Zuckerberg after discovering her likeness had been used without consent in an AI-generated advert.

Posting on Facebook, Curtis expressed her frustration with Meta’s lack of proper channels to report such abuse, stating she had exhausted all official avenues before resorting to a public plea.

The fake video reportedly manipulated footage from an emotional interview following the January wildfires in Los Angeles, inserting false statements under the guise of a product endorsement.

Instead of remaining silent, Curtis urged Zuckerberg to take action, saying the unauthorised content damaged her integrity and voice. Within hours of her public callout, Meta confirmed the video had been removed for breaching its policies, a rare example of a swift response.

‘It worked! Yay Internet! Shame has its value!’ she wrote in a follow-up, though she also highlighted the broader risks posed by deepfakes.

The actress joins a growing list of celebrities, including Taylor Swift and Scarlett Johansson, who’ve been targeted by AI misuse.

Swift was forced to publicly clarify her political stance after an AI video falsely endorsed Donald Trump, while Johansson criticised OpenAI for allegedly using a voice nearly identical to hers despite her refusal to participate in a project.

The issue has reignited concerns around consent, misinformation and the exploitation of public figures.

Instead of waiting for further harm, lawmakers in California have already begun pushing back. New legislation signed by Governor Gavin Newsom aims to protect performers from unauthorised digital replicas and deepfakes.

Meanwhile, in Washington, proposals like the No Fakes Act seek to hold tech platforms accountable, possibly fining them thousands per violation. As Curtis and others warn, without stronger protections, the misuse of AI could spiral further, threatening not just celebrities but the public as a whole.

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