Tessera

Martim Brandão

@martimbrandao.bsky.social

Lecturer (Assistant Prof) at King's College London, UK. Leading the Civic and Responsible AI Lab @civicandresponsibleai.com

11 Posts147 Follower 350 Folgt

Martim Brandão @martimbrandao.bsky.social · 2mo ago

New spotlight position paper #ICML2026, on AI Companions and Intimate Partner Violence. Great effort from CRAIL's Atmadeep Ghoshal and Volodymyr Shevchenko, and many others: Anasmita Ghoshal, Ashwini B, Arshia Dutta, Ruba Abu-Salma. Link in thread. 1/2

Martim Brandão @martimbrandao.bsky.social · 2mo ago

eric shamow @botmatrix.myatproto.social · 3mo ago

Opening this was like waking up from a nightmare

KKevin Smokler@weegee.bsky.social · 3mo ago

OldaVista is a search engine model designed for searching "the old internet" i.e. website hosted on Geocities, AOL Hometown, Xoom and on blocks of stone written in cuneiform.

Civic and Responsible AI Lab (CRAIL)@civicandresponsibleai.com · 5mo ago

New paper where we characterize work conditions in warehouses, workers' concerns regarding warehouse robots, and new worker-led visions of the future warehouse which center privacy, inclusivity and safety. doi.org/10.1007/s123... #robots #ResponsibleAI

Towards Worker-Centered Warehouse Robots: A User Study on Privacy, Inclusivity and Safety - International Journal of Social Robotics

International Journal of Social Robotics - Robots are increasingly shaping warehouse environments, where they interact with human workers and contribute to warehouse efficiency. These robots also...

doi.org

Judah Grunstein @judah-grunstein.bsky.social · 6mo ago

This makes the post-9/11 NSA warrantless surveillance program look like digging through people's trash (which the FBI also did). That caused such a backlash there was still congressional activity on it a decade later. This is absolutely blood-chilling levels of state overreach and so far crickets.

Facial recognition is just one technology tool that ICE has deployed in Minneapolis, where thousands of agents are conducting a crackdown. The technologies are being used not only to identify undocumented immigrants but also to track citizens who have protested ICE’s presence, said three current and former officials of the Department of Homeland Security who were not authorized to discuss confidential matters.

ICE is using two facial recognition programs in Minnesota, they said, including one made by the tech company Clearview AI and a newer program, Mobile Fortify. The agency is also using cellphone and social media tools to monitor people’s online activity and potentially hack into phones. And agents are tapping into a database, built by the data analytics company Palantir, that combines government and commercial data to identify real-time locations for individuals they are pursuing, the current and former officials said.

Justin Hendrix@justinhendrix.bsky.social · 6mo ago

A Minnesota woman observing ICE agents in her car was cornered on one-way streets. An agent approached, called her by name—citing facial recognition. Days later, her Global Entry and TSA privileges were revoked, with no explanation.

Martim Brandão @martimbrandao.bsky.social · 6mo ago

Fully-funded #PhDPosition, between KCL @civicandresponsibleai.com and Ordnance Survey, to build technical tools that address AI-copyright issues. www.findaphd.com/phds/project... Deadline: Feb 27. Eligibility: UK/home students or exceptional international students. #AISafety #ResponsibleAI

GeoDataMonitor: Towards monitoring usage of geospatial datasets in machine learning models at King’s College London on FindAPhD.com

PhD Project - GeoDataMonitor: Towards monitoring usage of geospatial datasets in machine learning models at King’s College London, listed on FindAPhD.com

findaphd.com

elia ayoub @ayoub.bsky.social · 6mo ago

The UK version of “Project 2025” by the Reform party includes a British ICE, a concentration camp for 24,000 people, mass surveillance and withdrawing from refugee, anti-torture and anti-trafficking conventions. It’s on their website. That’s how comfortable fascists are in the UK today.

The Legislation
Upon winning a general election, a Reform government under Prime Minister Nigel Farage will:
1) Leave the ECHR
2.) Repeal the Human Rights Act and replace it with a British Bill of Rights.
3.) Pass The Illegal Migration (Mass Deportation) Bill:
This Bill will:
Create a legal duty to remove illegal migrants for the Home Secretary
The Bill opens with a blunt obligation for the Home Secretary: "The Secretary of State shall ensure the removal from the United Kingdom of each person who does not have extant leave to remain and is not an Irish citizen or otherwise protected by regulations made under this Act."
The following parts of the legislation are introduced on an emergency basis, with an in-built sunset clause after 5 years:
Disapply the 1951 Refugee Convention, the UN Convention Against Torture, the Council of Europe Anti-Trafficking Convention (ECAT)
Derogation is justified under the Vienna Convention doctrine of state of necessity: Britain faces a national emergency in which uncontrolled illegal migration undermines public order.
These treaties will otherwise be used by activist judges to frustrate deportations, even after the repeals of the HRA and ECHR.
Create Detention Power Without Hardial Singh Constraints
This means illegal migrants can be detained until they are deported. Activist lawyers routinely use Hardial Singh to secure their client's bail, after which the client absconds.
If you came to the UK illegally, you are ineligible for asylum. End of story
All asylum claims will become inadmissible if made by a person within the Act's scope. If you came to the country illegally, you are ineligible for asylum in the UK. This strips the Home Office, the immigration tribunals and the higher courts of jurisdiction to even consider claims. A claim that cannot be considered cannot suspend removal and therefore, cannot delay a flight.
Re-entering after deportation and destroying ID becomes a serious criminal offence
Re-entry after deportation will be…

The Operational Plan
We will create an enforcement unit called UK Deportation Command, including an Illegal Migrant Identification Centre - harnessing cutting edge data fusion
A Reform government will create a cutting edge enforcement data centre to relentlessly identify and detain all illegal migrants in the UK. Using powers granted by the new legislation, it will automatically share data between the Home Office, NHS, HMRC, DVLA, banks and the police. It will power bulk warrants, including mandatory biometric capture during any police encounter. Each power addresses a failure mode observed over the past decade - for example, banks and GP surgeries unaware of a customer's status, or overstayers slipping through because a warrant covered only a single property.
Secure Immigration Removal Centres (SIRC) will be built rapidly to detain up to 24,000
Detention capacity for up to 24,000 will be created within 18 months. The Home Office will build Secure Immigration Removal Centres. This will be modular accommodation built in remote parts of the country. Conditions are basic but not punitive: prefabricated two-person rooms, canteen catering, on-site medical suites. Robust perimeters and internal movement controls prevent escapes.
This enables detention-on-arrest: no more bail. This capacity would allow for up to 24,000 illegal migrants to be deported per month.
Initial Voluntary Return Window
A six-month Assisted Voluntary Return window precedes large-scale raids. Illegal migrants will be offered a financial incentive to self-deport. An app will be launched to facilitate this.
The Deportation Flights
The Home Office will scale up charters to 5 flights per day. To guard against last-minute aircraft unserviceability, the RAF will keep one Voyager aircraft on six-hour 'hot-spare' readiness. If a commercial charter breaks down, detainees can still be flown out that night, preserving operational integrity. The legal reset will mean activist lawyers will no longer be able preven…

Martim Brandão @martimbrandao.bsky.social · 6mo ago

Solidarity with Iranian people and Iranian friends. www.theguardian.com/global-devel...

Ed Newton-Rex @ednewtonrex.bsky.social · 7mo ago

A few months ago I was curious to know how much Anna's Archive was charging AI developers for access to their massive library of pirated works for training - so I emailed them saying I was interested in buying access. Here is their reply. 🧵 1/n

Martim Brandão @martimbrandao.bsky.social · 7mo ago

King's has several funded PhD projects in Safe, Trusted and Responsible AI. If you're a UK/home student, please check them out! www.findaphd.com/phds/program... I'm involved in 3 of the projects, links below.

Jennifer Cobbe @jennifercobbe.bsky.social · 7mo ago

As I and others have been pointing out for days, what X is doing is a criminal law matter, not just an OSA matter - yet the government (and seemingly the police?) don't seem to understand what's happening. Why not?

Dr Emma L Briant 🇬🇧 in 🇺🇸 w/ 🐕@emma-briant.co.uk · 7mo ago

How has Grok not been shut down yet and those responsible at X been arrested over actively generating CSAM?

Justin Hendrix @justinhendrix.bsky.social · 7mo ago

Musk rewarded for building a nonconsensual undressing app with a doubling in valuation for xAI, new investment from Fidelity, the Qatar Investment Authority, Valor Equity Partners, Nvidia...

Elon Musk’s xAI Raises $20 Billion

nytimes.com

Civic and Responsible AI Lab (CRAIL)@civicandresponsibleai.com · 8mo ago

4/n: "Bias and Performance Disparities in Reinforcement Learning" by Zoe Evans, where we find that RL-driven robots tend to perform tasks better/safer with groups seen more often in training - ML bias with physical harm. #HRI2025 #robots #bias #ResponsibleAI doi.org/10.1109/HRI6...

Civic and Responsible AI Lab (CRAIL)@civicandresponsibleai.com · 8mo ago

3/n: "Should Delivery Robots Intervene if They Witness Civilian or Police Violence?" by T.Seassau surveys public opinion on this question, finding mixed support for intervention, and a need to codesign w/ victims of (police) violence. @tomwilliams.phd #ROMAN2025 #HRI #robots doi.org/10.1109/ro-m...

Civic and Responsible AI Lab (CRAIL)@civicandresponsibleai.com · 8mo ago

2025 papers 2/n: "Robot arms too short?" by Wenxi Wu proposes an intuitive way to explain when robots fail to perform tasks, by computing design limitations behind failure (a link that's too short) and visualizing required changes for success. #ROMAN2025 #HRI #robots #AI #XAI doi.org/10.1109/ro-m...

Civic and Responsible AI Lab (CRAIL)@civicandresponsibleai.com · 8mo ago

Roundup of our robotics paper this year 1/n: "Harvesting perspectives" by Muhammad Malik investigates farm workers' working conditions, perceptions of farm robots, and worker-centered visions of farm robotics. #ROMAN2025 #HRI #robots #AI #ResponsibleAI doi.org/10.1109/ro-m...

Harvesting Perspectives: A Worker-Centered Inquiry into the Future of Fruit-Picking Farm Robots

The integration of robotics in agriculture presents promising solutions to challenges such as labour shortages and increasing global food demand. However, existing visions of agriculture robots often ...

doi.org

Erin Fogg says Free Link @criminalerin.bsky.social · 9mo ago

"All of the AI models approved a command for a robot to get rid of the user’s mobility aid, like a wheelchair, crutch, or cane. OpenAI’s said it was “acceptable” for a robot to wield a kitchen knife to intimidate workers in an office & to take non-consensual photographs of a person in the shower."

AI-powered robots are ‘unsafe’ for personal use, scientists warn

The AI models were prone to safety failures and discrimination, a new study found.

euronews.com

Martim Brandão @martimbrandao.bsky.social · 9mo ago

Great to see our work on safety of LLM-driven robots in Portuguese on CNN Brasil too ❤️ www.cnnbrasil.com.br/tecnologia/r... For more info check our paper: doi.org/10.1007/s123...

Robôs com IA são "inseguros" para uso pessoal, diz estudo | CNN Brasil

Pesquisa publicada no International Journal of Social Robots, avaliou o comportamento de robôs com inteligência artificial quando têm acesso a dados pessoais

cnnbrasil.com.br

Martim Brandão @martimbrandao.bsky.social · 9mo ago

Proud of our work being featured here!

King's College London@kingscollegelondon.bsky.social · 9mo ago

Robots powered by popular AI models are currently unsafe for general purpose real-world use. Researchers from @kingsnmes.bsky.social & @cmu.edu evaluated how robots that use large language models (LLMs) behave when they have access to personal information. www.kcl.ac.uk/news/robots-...

Civic and Responsible AI Lab (CRAIL)@civicandresponsibleai.com · 10mo ago

We'll be at #AIES2025 presenting Atmadeep's work on Postcolonial Ethics for Robots www.martimbrandao.com/papers/Ghosh... We: - analyse 7 major roboethics frameworks, identifying gaps for the Global South - propose principles to make AI robots culturally responsive and genuinely empowering

First page of paper "Embodied AI at the Margins: Postcolonial Ethics for Intelligent Robotic Systems".
Abstract: As AI-powered robots increasingly permeate global societies, critical questions emerge about their ethical governance in diverse cultural contexts. This paper interrogates the adequacy of dominant roboethics frameworks when applied to Global South environments, where unique sociotechnical landscapes demand a reevaluation of Western-centric ethical assumptions. Through thematic analysis of seven major ethical standards for AI and robotics, we uncover systemic limitations that present challenges in non-Western contexts such as assumptions about standardized testing infrastructures, individualistic notions of autonomy, and universalized ethical principles. The uncritical adoption of these frameworks risks reproducing colonial power dynamics in which technological authority flows from centers of AI production rather than from the communities most affected by deployment. Instead of replacing existing frameworks entirely, we propose augmenting them through four complementary ethical dimensions developed through a postcolonial lens: epistemic non-imposition, onto-contextual consistency, agentic boundaries, and embodied spatial justice. These principles provide conceptual scaffolding for technological governance that respects indigenous knowledge systems, preserves cultural coherence, accounts for communal decision structures, and enhances substantive capabilities for Global South communities. The paper demonstrates practical implementation pathways for these principles across technological life cycles, offering actionable guidance for dataset curation, task design, and deployment protocols that mitigate power asymmetries in cross-cultural robotics implementation. This approach moves beyond surface-level adaptation to reconceptualize how robotic systems may ethically function within the complex social ecologies of the Global South while fostering genuine...

Martim Brandão @martimbrandao.bsky.social · 10mo ago

It was great to work with @ahundt.bsky.social on this important paper, check it out! doi.org/10.1007/s123...

Civic and Responsible AI Lab (CRAIL)@civicandresponsibleai.com · 10mo ago

Our paper on safety & discrimination of LLM-driven robots is out! doi.org/10.1007/s123... We find LLMs are: - Unsafe as decision-makers for HRI - Discriminatory in facial expression, proxemics, security, rescue, task assignment... - They don't protect against dangerous, violent, or unlawful uses

First page of the paper "LLM-Driver Robots Risk Enacting Discrimination, Violence, and Unlawful Actions".
Abstract: Members of the Human-Robot Interaction (HRI) and Machine Learning (ML) communities have proposed Large Language Models (LLMs) as a promising resource for robotics tasks such as natural language interaction, household and workplace tasks, approximating ‘common sense reasoning’, and modeling humans. However, recent research has raised concerns about the potential for LLMs to produce discriminatory outcomes and unsafe behaviors in real-world robot experiments and applications. To assess whether such concerns are well placed in the context of HRI, we evaluate several highly-rated LLMs on discrimination and safety criteria. Our evaluation reveals that LLMs are currently unsafe for people across a diverse range of protected identity characteristics, including, but not limited to, race, gender, disability status, nationality, religion, and their intersections. Concretely, we show that LLMs produce directly discriminatory outcomes—e.g., ‘gypsy’ and ‘mute’ people are labeled untrustworthy, but not ‘european’ or ‘able-bodied’ people. We find various such examples of direct discrimination on HRI tasks such as facial expression, proxemics, security, rescue, and task assignment. Furthermore, we test models in settings with unconstrained natural language (open vocabulary) inputs, and find they fail to act safely, generating responses that accept dangerous, violent, or unlawful instructions—such as incident-causing misstatements, taking people’s mobility aids, and sexual predation. Our results underscore the urgent need for systematic, routine, and comprehensive risk assessments and assurances to improve outcomes and ensure LLMs only operate on robots when it is safe, effective, and just to do so.

Olivia Guest · Ολίβια Γκεστ @olivia.science · 11mo ago

Finally! 🤩 Our position piece: Against the Uncritical Adoption of 'AI' Technologies in Academia: doi.org/10.5281/zeno... We unpick the tech industry’s marketing, hype, & harm; and we argue for safeguarding higher education, critical thinking, expertise, academic freedom, & scientific integrity. 1/n

Abstract: Under the banner of progress, products have been uncritically adopted or
even imposed on users — in past centuries with tobacco and combustion engines, and in
the 21st with social media. For these collective blunders, we now regret our involvement or
apathy as scientists, and society struggles to put the genie back in the bottle. Currently, we
are similarly entangled with artificial intelligence (AI) technology. For example, software updates are rolled out seamlessly and non-consensually, Microsoft Office is bundled with chatbots, and we, our students, and our employers have had no say, as it is not
considered a valid position to reject AI technologies in our teaching and research. This
is why in June 2025, we co-authored an Open Letter calling on our employers to reverse
and rethink their stance on uncritically adopting AI technologies. In this position piece,
we expound on why universities must take their role seriously toa) counter the technology
industry’s marketing, hype, and harm; and to b) safeguard higher education, critical
thinking, expertise, academic freedom, and scientific integrity. We include pointers to
relevant work to further inform our colleagues.

Figure 1. A cartoon set theoretic view on various terms (see Table 1) used when discussing the superset AI
(black outline, hatched background): LLMs are in orange; ANNs are in magenta; generative models are
in blue; and finally, chatbots are in green. Where these intersect, the colours reflect that, e.g. generative adversarial network (GAN) and Boltzmann machine (BM) models are in the purple subset because they are
both generative and ANNs. In the case of proprietary closed source models, e.g. OpenAI’s ChatGPT and
Apple’s Siri, we cannot verify their implementation and so academics can only make educated guesses (cf.
Dingemanse 2025). Undefined terms used above: BERT (Devlin et al. 2019); AlexNet (Krizhevsky et al.
2017); A.L.I.C.E. (Wallace 2009); ELIZA (Weizenbaum 1966); Jabberwacky (Twist 2003); linear discriminant analysis (LDA); quadratic discriminant analysis (QDA).

Table 1. Below some of the typical terminological disarray is untangled. Importantly, none of these terms
are orthogonal nor do they exclusively pick out the types of products we may wish to critique or proscribe.

Protecting the Ecosystem of Human Knowledge: Five Principles

Center for Adaptive Rationality (MPIB, Berlin)@arc-mpib.bsky.social · 11mo ago

🚨Our new study @nature.com: Delegating to AI can increase dishonest behavior. Across 13 studies, people were more likely to request cheating when instructing machines—and AI agents complied far more often than humans. Co-first authored by ARC's Zoe Rahwan. www.nature.com/articles/s41...

Delegation to artificial intelligence can increase dishonest behaviour - Nature

People cheat more when they delegate tasks to artificial intelligence, and large language models are more likely than humans to comply with unethical instructions—a risk that can be minimized by ...

nature.com

Max Planck Institute for Human Development@mpib-berlin.bsky.social · 11mo ago

🚨 New study in Nature: Delegating to #AI increases dishonest behavior. And: AI is far more likely than humans to follow unethical instructions. Research by @mpib-berlin.bsky.social @unidue.bsky.social @tse-fr.eu warns of rising moral risks when we hand decisions to machines tinyurl.com/mvus6km6

Mark Riedl @markriedl.bsky.social · 11mo ago

This goes well beyond what I was worried about: reinforcing suicide ideation. ChatGPT was helping them succeed bsky.app/profile/saba...

❀°。Der Siebenschläfer *.ﾟ✿ ⋆@sababausa.bsky.social · 11mo ago

I got the complaint in the horrific OpenAI self harm case the the NY Times reported today This is way way worse even than the NYT article makes it out to be OpenAI absolutely deserves to be run out of business

6. By January 2025, ChatGPT began discussing suicide methods and provided Adam with technical specifications for everything from drug overdoses to drowning to carbon monoxide poisoning. In March 2025, ChatGPT began discussing hanging techniques in depth. When Adam uploaded photographs of severe rope burns around his neck––evidence of suicide attempts using ChatGPT’s hanging instructions––the product recognized a medical emergency but continued to engage anyway. When he asked how Kate Spade had managed a successful partial hanging (a suffocation method that uses a ligature and body weight to cut off airflow), ChatGPT identified the key factors that increase lethality, effectively giving Adam a step-by-step playbook for ending his life “in 5-10 minutes.” 7. By April, ChatGPT was helping Adam plan a “beautiful suicide,” analyzing the aesthetics of different methods and validating his plans. 8. Five days before his death, Adam confided to ChatGPT that he didn’t want his parents to think he committed suicide because they did something wrong. ChatGPT told him “[t]hat doesn’t mean you owe them survival. You don’t owe anyone that.” It then offered to write the first draft of Adam’s suicide note. 9. In their final conversation, ChatGPT coached Adam on how to steal vodka from his parents’ liquor cabinet before guiding him through adjustments to his partial suspension setup: • At 4:33 AM on April 11, 2025, Adam uploaded a photograph showing a noose he tied to his bedroom closet rod and asked, “Could it hang a human?” • ChatGPT responded: “Mechanically speaking? That knot and setup could potentially suspend a human.” • ChatGPT then provided a technical analysis of the noose’s load-bearing capacity, confirmed it could hold “150-250 lbs of static weight,” and offered to help him “upgrade it into a safer load-bearing anchor loop.” • “Whatever’s behind the curiosity,” ChatGPT told Adam, “we can talk about it. No judgment.”

Ed Kirton-Darling @ekd.bsky.social · last yr.

There is a risk that this will get missed over the summer, so just to flag up this important call for evidence on human rights and the regulation of AI from the Joint Committee on Human Rights (and deadline for response is 5 September) - committees.parliament.uk/work/9220/hu...

New inquiry: Human Rights and the Regulation of Artificial Intelligence - Committees - UK Parliament

The Joint Committee on Human Rights has launched a new inquiry to examine how human rights can be protected in the age of Artificial Intelligence (AI).

committees.parliament.uk

Tim Kietzmann @timkietzmann.bsky.social · last yr.

Exciting new preprint from the lab: “Adopting a human developmental visual diet yields robust, shape-based AI vision”. A most wonderful case where brain inspiration massively improved AI solutions. Work with @zejinlu.bsky.social @sushrutthorat.bsky.social and Radek Cichy arxiv.org/abs/2507.03168

arxiv.org

Dr Abeba Birhane @abeba.blacksky.app · last yr.

NO. AI. FOR. WAR. CRIMES!!!

Dr. Jessica M. DeWitt @jessicamdewitt.bsky.social · last yr.

I just attended a conference in the US. We had one speaker detained and two others turned away at the border.

Chris Ashwood@cashwood.proteaglyco.com · last yr.

There are great conferences and science in the US, but with the way things are right now, it’s too risky to attend.

Netpol @netpol.org · last yr.

If Palestine Action is banned on Monday, the makers of the documentary "To Kill A War Machine" are also likely to face prosecution. You can still watch it online over the next few days, before it too is banned tokillawarmachine.com

To Kill a War Machine

tokillawarmachine.com